A method and system for alveolar surgical procedure risk assessment

Abstract

The application relates to the technical field of medical data processing, and discloses a dental surgical operation risk assessment method and system, which comprises the following steps: acquiring oral cavity parameters and dental surgical operation type identifiers as knowledge graph nodes, taking historical risk parameters, historical tooth risk parameters and historical health risk parameters as knowledge graph edge weights, obtaining a dental surgical operation risk correlation matrix according to a dental surgical operation knowledge graph, obtaining a deviation probability based on user oral cavity parameters and historical oral cavity parameters, solving the dental surgical operation risk correlation matrix based on the user oral cavity parameters and the deviation probability, obtaining an initial operation risk assessment result, acquiring operation type risk factors based on the dental surgical operation type identifiers, and fusing the operation type risk factors and the initial operation risk assessment result to obtain a dental surgical operation risk assessment result. The application solves the technical problem of insufficient accuracy of dental surgical operation risk assessment in the prior art.

Description

Technical Field

[0001] This invention relates to the field of medical data processing technology, and in particular to a method and system for assessing the risk of alveolar surgery. Background Technology

[0002] Alveolar surgery is a common oral surgical treatment, which has certain invasiveness and potential risks. Such surgery usually requires a comprehensive judgment based on the patient's tooth type, alveolar bone condition and the surgical procedure to be performed. If the preoperative assessment is insufficient, it may lead to risk events such as bleeding, infection, bone wall destruction, and damage to adjacent anatomical structures, affecting the safety of the surgery and the stability of the prognosis. In existing technologies, preoperative risk assessment often relies on physician experience and rule-based judgment, or qualitative analysis of individual conditions based on imaging and medical records. While it's possible to extract tooth categories from electronic medical records or imaging reports and use CBCT images to calculate bone mineral density indicators and measure alveolar bone height for assessment, these methods often have the following shortcomings: First, risk factors are diverse and significantly interconnected, making it difficult for traditional threshold judgments to characterize the strength of the association and propagation path between risk factors. Second, many assessment conclusions lack sufficient probabilistic quantitative comparison with historical case data from patients undergoing the same procedure, tooth category, and similar oral health conditions, leading to insufficient objectivity and interpretability of the assessment results. Furthermore, in real-world scenarios, oral parameters from different patients undergoing the same procedure may deviate significantly from typical historical distributions. Without characterizing the degree of difference between individual parameters and historical parameter sets for the same procedure, assessment results are prone to instability or insufficient generalization when patients are atypical samples. Additionally, existing assessments often struggle to quantify risk factors such as the complexity of the procedure itself and integrate them with the initial assessment results to obtain a more comprehensive risk output that aligns with clinical decision-making. Traditional surgical risk assessment relies heavily on doctors' clinical experience, which is subject to strong subjectivity and inconsistent assessment standards. It is also difficult to fully quantify the intrinsic relationship between individual differences and surgical risks, which can easily lead to inaccurate risk predictions. This can affect the precise optimization of surgical plans and the effective prevention and control of complications. Therefore, there is an urgent need for a method and system for alveolar surgery risk assessment. Summary of the Invention

[0003] The purpose of this invention is to overcome one or more of the above-mentioned existing technical problems and provide a method and system for assessing the risks of alveolar surgery.

[0004] To achieve the above objectives, the present invention provides a method for assessing the risk of alveolar surgery, comprising: Obtain oral parameters and alveolar surgical procedure identifiers, and use these parameters and identifiers as nodes in the alveolar surgical knowledge graph. Historical surgical risk parameters, historical tooth risk parameters, and historical health risk parameters are obtained based on alveolar surgical procedure identifiers, tooth categories, and oral health parameters, and are used as edge weights for the alveolar surgical knowledge graph. Feature embedding calculations are performed based on nodes and edge weights of the alveolar surgery knowledge graph to construct the alveolar surgery knowledge graph and obtain the alveolar surgery risk association matrix. Collect user oral parameters and obtain historical oral parameters corresponding to alveolar surgical procedure identifiers; calculate the probability of alveolar surgical detection deviation based on user oral parameters and historical oral parameters. The risk correlation matrix of alveolar surgery is solved based on user oral parameters and the probability of deviation in alveolar surgery detection to obtain the initial risk assessment results of the surgery. Based on the surgical procedure identifier, surgical procedure risk factors are obtained, and the surgical procedure risk factors are integrated with the initial surgical risk assessment results to obtain the surgical risk assessment results of alveolar surgery.

[0005] According to one aspect of the invention, oral parameters include at least tooth category and oral health parameters; Based on oral parameters and alveolar surgical procedure identifiers as nodes in the alveolar surgical knowledge graph, the tooth category, bone mineral density value, alveolar bone height value and alveolar surgical procedure identifier in oral health parameters are each treated as an independent node; Tooth categories are extracted from electronic medical records or imaging reports, bone mineral density values ​​are calculated using the average grayscale value of CBCT image pixels, and alveolar bone height values ​​are calculated using the vertical distance between two points on the CBCT image.

[0006] According to one aspect of the present invention, the probability of occurrence of risk events in historical cases of the same alveolar surgical procedure is statistically analyzed based on the identification of the surgical procedure and used as a historical surgical risk parameter; Historical dental risk parameters are based on the probability of occurrence of risk events corresponding to different tooth categories. The probability of occurrence of risk events is statistically analyzed based on oral health parameters and used as historical health risk parameters; The edge weights between nodes in the alveolar surgery knowledge graph are calculated based on historical surgical risk parameters, historical dental risk parameters, and historical health risk parameters, thereby representing the probability of risk association between nodes.

[0007] According to one aspect of the present invention, a graph neural network model is used to perform feature embedding calculations on nodes in an alveolar surgery knowledge graph to generate initial embedding vectors for each node. The embedding vector of the target node is updated by weighting and aggregating the embedding vectors of neighboring nodes connected to the target node based on edge weights. The updated node embedding vectors are combined as row vectors to form a node embedding matrix. The product of the node embedding matrix and its transpose is calculated to obtain the alveolar surgery risk association matrix, which represents the association strength of any node pair in the risk feature space.

[0008] According to one aspect of the present invention, a set of historical oral parameters for the same surgical procedure is obtained based on alveolar surgical procedure identifiers; The mean and standard deviation of numerical parameters in the historical oral parameter set are calculated, and the probability of numerical deviation is determined based on the degree of deviation between the user's numerical oral parameters and the mean. The historical frequency percentage of the classification parameters is statistically analyzed, and the probability of classification deviation is determined by the complementary value of the percentage. The numerical deviation probability and the categorical deviation probability are summarized to obtain the alveolar surgical detection deviation probability, which is used to represent the degree of difference between the user's oral parameters and the historical oral parameters of the same procedure.

[0009] According to one aspect of the present invention, based on user oral parameters and alveolar surgical procedure identifiers, a node-related submatrix of the corresponding node is extracted from the alveolar surgical risk association matrix; The risk association submatrix is ​​obtained by correcting the node-related submatrix based on the probability of deviation in alveolar surgery detection; The typicality of a surgical procedure is obtained by obtaining the probability of occurrence of the risk correlation submatrix within the risk correlation matrix of alveolar surgery. A user oral cavity parameter vector is constructed based on the user's oral cavity parameters, and a matrix operation is performed between the user oral cavity parameter vector and the risk correlation submatrix to obtain the user risk vector. The user risk assessment results are obtained based on the user risk vector, and then the user risk assessment results are corrected based on the typicality of the surgical procedure to obtain the initial surgical risk assessment results.

[0010] According to one aspect of the present invention, the surgical procedure risk factor includes at least a numerical coefficient corresponding to the surgical complexity, and a comprehensive surgical procedure risk factor is obtained based on the numerical coefficient corresponding to the surgical complexity and other surgical procedure risk factors. When there are multiple surgical procedure risk factors, a weighted average of the multiple surgical procedure risk factors is taken to obtain a comprehensive surgical procedure risk factor. The risk assessment results of alveolar surgery are obtained by integrating the comprehensive surgical procedure risk factors with the initial surgical risk assessment results.

[0011] To achieve the above objectives, the present invention provides a risk assessment system for alveolar surgery, comprising: Alveolar surgery knowledge graph edge node construction module: Obtain oral parameters and alveolar surgery procedure identifiers, and use oral parameters and alveolar surgery procedure identifiers as nodes of alveolar surgery knowledge graph; Alveolar surgery knowledge graph edge weight construction module: Based on alveolar surgery procedure identifier, tooth category and oral health parameters, historical surgical risk parameters, historical tooth risk parameters and historical health risk parameters are obtained respectively, and used as edge weights of alveolar surgery knowledge graph; Alveolar surgery risk association matrix acquisition module: Based on the nodes and edge weights of the alveolar surgery knowledge graph, feature embedding calculation is performed to construct the alveolar surgery knowledge graph and obtain the alveolar surgery risk association matrix; Alveolar surgery detection deviation probability generation module: Collects user oral parameters and obtains historical oral parameters corresponding to alveolar surgery procedure identifiers, and generates the alveolar surgery detection deviation probability based on user oral parameters and historical oral parameters; The module for obtaining the initial surgical risk assessment results solves the correlation matrix of alveolar surgery risk based on the user's oral parameters and the probability of deviation in alveolar surgery detection, thus obtaining the initial surgical risk assessment results. Alveolar surgery risk assessment result acquisition module: Based on the alveolar surgery procedure identifier, the procedure risk factors are obtained, and the procedure risk factors are integrated with the initial surgical risk assessment results to obtain the alveolar surgery risk assessment results.

[0012] To achieve the above objectives, the present invention provides an electronic device, including a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the above-described method for risk assessment of alveolar surgery.

[0013] To achieve the above objectives, the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described method for risk assessment of alveolar surgery.

[0014] Based on this, the beneficial effects of the present invention are as follows: Firstly, by acquiring oral parameters and alveolar surgical procedure identifiers as nodes in the alveolar surgical knowledge graph, the invention integrates individual patient oral characteristics and surgical type characteristics that influence surgical risk, providing basic data units for the construction of the alveolar surgical knowledge graph. Secondly, by performing feature embedding calculations on the alveolar surgical graph to construct the alveolar surgical knowledge graph and obtaining the alveolar surgical risk association matrix, the invention achieves unified quantification and structured integration of heterogeneous node features in the alveolar surgical graph, accurately characterizing the strength of deep risk associations between nodes, and providing a basis for subsequent risk assessment results based on individual user parameters. Based on the computational model, the user's oral parameters are collected. Based on these parameters, the risk correlation matrix of alveolar surgery is solved to obtain the initial risk assessment results. This achieves a precise fit between the general alveolar surgery risk correlation matrix and the user's individual oral characteristics, quantifying the user's basic surgical risk based on their own oral parameters. Finally, based on the alveolar surgery procedure identifier, the procedure risk factor is obtained. Combined with the initial risk assessment results, the alveolar surgery risk assessment results are obtained. This achieves the fusion of the inherent risk of the procedure and the user's individual basic risk, making up for the limitations of a single assessment and obtaining a comprehensive and accurate alveolar surgery risk assessment result. By converting oral parameters and alveolar surgical procedure identifiers into knowledge graph nodes, a risk association matrix for alveolar surgery is constructed through feature embedding calculations. The initial risk assessment results are then solved by combining individual user oral characteristics, and the final assessment is completed by incorporating procedure risk factors. This effectively overcomes the shortcomings of traditional assessments, which rely on experience, are highly subjective, and lack standardized criteria. It achieves a systematic, objective, and comprehensive quantification of alveolar surgical risks, accurately links individual differences with inherent risks of the procedure, improves the accuracy of risk prediction, and provides reliable support for surgical plan optimization and surgical safety assurance, thus meeting the needs of precision medicine development. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating a method for assessing the risk of alveolar surgery according to an exemplary embodiment; Figure 2 This is a flowchart illustrating a risk assessment system for alveolar surgery according to an exemplary embodiment. Detailed Implementation

[0016] The invention will now be discussed with reference to exemplary embodiments. It should be understood that the described embodiments are merely intended to enable those skilled in the art to better understand and thus implement the invention, and are not intended to imply any limitation on the scope of the invention.

[0017] As used herein, the term “comprising” and its variations are to be interpreted as open-ended terms meaning “including but not limited to”. The term “based on” is to be interpreted as “at least partially based on”, and the terms “one embodiment” and “an embodiment” are to be interpreted as “at least one embodiment”.

[0018] According to one embodiment of the present invention, Figure 1 This is a flowchart illustrating a method for assessing the risk of alveolar surgery according to an exemplary embodiment, such as... Figure 1 As shown, to achieve the above objectives, the present invention provides a method for assessing the risk of alveolar surgery, comprising: Obtain oral parameters and alveolar surgical procedure identifiers, and use these parameters and identifiers as nodes in the alveolar surgical knowledge graph. Historical surgical risk parameters, historical tooth risk parameters, and historical health risk parameters are obtained based on alveolar surgical procedure identifiers, tooth categories, and oral health parameters, and are used as edge weights for the alveolar surgical knowledge graph. Feature embedding calculations are performed based on nodes and edge weights of the alveolar surgery knowledge graph to construct the alveolar surgery knowledge graph and obtain the alveolar surgery risk association matrix. Collect user oral parameters and obtain historical oral parameters corresponding to alveolar surgery procedure identifiers. Based on user oral parameters and historical oral parameters, obtain the probability of alveolar surgery detection deviation. The risk correlation matrix of alveolar surgery is solved based on user oral parameters and the probability of deviation in alveolar surgery detection to obtain the initial risk assessment results of the surgery. Based on the surgical procedure identifier, surgical procedure risk factors are obtained, and the surgical procedure risk factors are integrated with the initial surgical risk assessment results to obtain the surgical risk assessment results of alveolar surgery.

[0019] According to one embodiment of the present invention, oral parameters include at least tooth category and oral health parameters; Based on oral parameters and alveolar surgical procedure identifiers as nodes in the alveolar surgical knowledge graph, the tooth category, bone mineral density value, alveolar bone height value and alveolar surgical procedure identifier in oral health parameters are each treated as an independent node; Tooth categories are extracted from electronic medical records or imaging reports, bone mineral density values ​​are calculated using the average grayscale value of CBCT image pixels, and alveolar bone height values ​​are calculated using the vertical distance between two points on the CBCT image.

[0020] According to one embodiment of the present invention, the probability of occurrence of risk events in historical cases of the same alveolar surgical procedure is statistically analyzed based on the identification of the surgical procedure and used as a historical surgical risk parameter; Historical dental risk parameters are based on the probability of occurrence of risk events corresponding to different tooth categories. The probability of occurrence of risk events is statistically analyzed based on oral health parameters and used as historical health risk parameters; The edge weights between nodes in the alveolar surgery knowledge graph are calculated based on historical surgical risk parameters, historical dental risk parameters, and historical health risk parameters, thereby representing the probability of risk association between nodes.

[0021] According to one embodiment of the present invention, a graph neural network model is used to perform feature embedding calculation on nodes in the alveolar surgery knowledge graph to generate an initial embedding vector for each node. The embedding vector of the target node is updated by weighting and aggregating the embedding vectors of neighboring nodes connected to the target node based on edge weights. The updated node embedding vectors are combined as row vectors to form a node embedding matrix. The product of the node embedding matrix and its transpose is calculated to obtain the alveolar surgery risk association matrix, which represents the association strength of any node pair in the risk feature space.

[0022] According to one embodiment of the present invention, a set of historical oral parameters for the same surgical procedure is obtained based on alveolar surgical procedure identifiers; The mean and standard deviation of numerical parameters in the historical oral parameter set are calculated, and the probability of numerical deviation is determined based on the degree of deviation between the user's numerical oral parameters and the mean. The historical frequency percentage of the classification parameters is statistically analyzed, and the probability of classification deviation is determined by the complementary value of the percentage. The numerical deviation probability and the categorical deviation probability are summarized to obtain the alveolar surgical detection deviation probability, which is used to represent the degree of difference between the user's oral parameters and the historical oral parameters of the same procedure.

[0023] According to one embodiment of the present invention, based on user oral parameters and alveolar surgery procedure identifiers, a node-related submatrix of the corresponding node is extracted from the alveolar surgery risk association matrix; The risk association submatrix is ​​obtained by correcting the node-related submatrix based on the probability of deviation in alveolar surgery detection; The typicality of a surgical procedure is obtained by obtaining the probability of occurrence of the risk correlation submatrix within the risk correlation matrix of alveolar surgery. A user oral cavity parameter vector is constructed based on the user's oral cavity parameters, and a matrix operation is performed between the user oral cavity parameter vector and the risk correlation submatrix to obtain the user risk vector. The user risk assessment results are obtained based on the user risk vector, and then the user risk assessment results are corrected based on the typicality of the surgical procedure to obtain the initial surgical risk assessment results.

[0024] According to one embodiment of the present invention, the surgical procedure risk factor includes at least a numerical coefficient corresponding to the surgical complexity, and a comprehensive surgical procedure risk factor is obtained based on the numerical coefficient corresponding to the surgical complexity and other surgical procedure risk factors. When there are multiple surgical procedure risk factors, a weighted average of the multiple surgical procedure risk factors is taken to obtain a comprehensive surgical procedure risk factor. The risk assessment results of alveolar surgery are obtained by integrating the comprehensive surgical procedure risk factors with the initial surgical risk assessment results.

[0025] Furthermore, to achieve the aforementioned objectives, this invention also provides a risk assessment system for alveolar surgery. Figure 2 This is a flowchart illustrating a risk assessment system for alveolar surgery according to an exemplary embodiment, such as... Figure 2 As shown, the alveolar surgical risk assessment system of the present invention includes: Alveolar surgery knowledge graph edge node construction module: Obtain oral parameters and alveolar surgery procedure identifiers, and use oral parameters and alveolar surgery procedure identifiers as nodes of alveolar surgery knowledge graph; Alveolar surgery knowledge graph edge weight construction module: Based on alveolar surgery procedure identifier, tooth category and oral health parameters, historical surgical risk parameters, historical tooth risk parameters and historical health risk parameters are obtained respectively, and used as edge weights of alveolar surgery knowledge graph; Alveolar surgery risk association matrix acquisition module: Based on the nodes and edge weights of the alveolar surgery knowledge graph, feature embedding calculation is performed to construct the alveolar surgery knowledge graph and obtain the alveolar surgery risk association matrix; Alveolar surgery detection deviation probability generation module: Collects user oral parameters and obtains historical oral parameters corresponding to alveolar surgery procedure identifiers, and generates the alveolar surgery detection deviation probability based on user oral parameters and historical oral parameters; The module for obtaining the initial surgical risk assessment results solves the correlation matrix of alveolar surgery risk based on the user's oral parameters and the probability of deviation in alveolar surgery detection, thus obtaining the initial surgical risk assessment results. Alveolar surgery risk assessment result acquisition module: Based on the alveolar surgery procedure identifier, the procedure risk factors are obtained, and the procedure risk factors are integrated with the initial surgical risk assessment results to obtain the alveolar surgery risk assessment results.

[0026] Specifically, the node acquisition module is used to acquire oral parameters and alveolar surgical procedure identifiers as nodes in the alveolar surgical knowledge graph.

[0027] The risks of alveolar surgery are influenced by both the patient's individual oral characteristics and the type of surgery. Traditional risk assessments rely heavily on the doctor's experience and lack systematic integration and quantitative correlation of multi-dimensional key data.

[0028] To address the aforementioned issues, this application obtains oral parameters and alveolar surgical procedure identifiers as nodes in an alveolar surgical knowledge graph.

[0029] Specifically, the node acquisition module includes: The oral parameter acquisition unit is used to acquire oral parameters, including tooth category and oral health parameters. The identification acquisition unit is used to acquire alveolar surgical procedure identification. Node building units are used to identify alveolar surgical procedures using oral parameters and techniques, serving as nodes in the alveolar surgical knowledge graph.

[0030] This application first obtains oral parameters, which are various indicators that reflect the oral condition and are extracted or calculated from the patient's medical data. The categories of oral parameters include tooth type and oral health parameters: tooth type refers to the specific type of the patient's tooth to be operated on, which can be obtained by directly extracting text information from the patient's electronic medical record or imaging report; oral health parameters are quantitative indicators that reflect the health status of oral tissues and can be obtained through specific detection or calculation methods.

[0031] For example, the tooth category of the patient's tooth to be operated on can be extracted from the patient's electronic medical record as the mandibular first molar; at the same time, the bone mineral density value can be calculated from the average gray value of the CBCT image pixels, and the alveolar bone height value can be calculated from the vertical distance between two points on the CBCT image, and the bone mineral density value and the alveolar bone height value can be used as oral health parameters.

[0032] Secondly, obtain the alveolar surgical procedure identifier. Specifically, the alveolar surgical procedure identifier is a standardized code or text identifier used to distinguish different types of alveolar surgical procedures. The alveolar surgical procedure identifier is a categorical variable, requiring no complex calculations, and can be directly read from the hospital's surgical planning system. The surgical planning system is a professional medical system used by the hospital to enter and manage patient surgical plans, which stores the standardized procedure code and corresponding name for each patient's scheduled surgery.

[0033] For example, if the patient is to undergo a complex tooth extraction, the alveolar surgical procedure identifier obtained from the surgical planning system is D7210, which corresponds to the surgical procedure name of complex tooth extraction.

[0034] Finally, oral parameters and alveolar surgical procedure identifiers were used as nodes in the alveolar surgery knowledge graph. Specifically, since oral parameters reflect the individual oral characteristics of patients and alveolar surgical procedure identifiers reflect the type of surgery, the two together constitute the core dimensions affecting surgical risk. Therefore, these two types of data were directly used as nodes in the alveolar surgery knowledge graph.

[0035] For example, if the obtained oral parameters are the mandibular first molar, bone density value of 256, alveolar bone height value of 12mm, and alveolar surgery procedure identifier of D7210, these four pieces of information are respectively used as four independent alveolar surgery knowledge graph nodes.

[0036] In summary, compared to existing technologies, this application obtains oral parameters and alveolar surgical procedure identifiers as nodes in the alveolar surgical knowledge graph. This integrates individual patient oral characteristics and surgical type characteristics that influence surgical risk, providing fundamental data units for the construction of the alveolar surgical knowledge graph.

[0037] The graph construction module is used to perform feature embedding calculations on alveolar surgery atlases, construct alveolar surgery knowledge graphs, and obtain alveolar surgery risk association matrices.

[0038] Because the nodes of the alveolar surgery knowledge graph contain heterogeneous features such as text and numerical values, and the edge weights are risk association probabilities, these discrete and heterogeneous data cannot directly quantify the deep risk associations between nodes, and traditional methods are difficult to integrate such data and accurately characterize the strength of risk associations between nodes.

[0039] To address the aforementioned issues, this application performs feature embedding calculations on the alveolar surgery atlas to construct an alveolar surgery knowledge graph and obtain an alveolar surgery risk association matrix.

[0040] Specifically, the map construction module includes: The historical surgical risk parameter acquisition unit is used to acquire historical surgical risk parameters based on alveolar surgical procedure identification. The historical dental risk parameter acquisition unit is used to acquire historical dental risk parameters based on dental category. The historical health risk parameter acquisition unit is used to acquire historical health risk parameters based on oral health parameters. The edge weight calculation unit is used to obtain the edge weights of the alveolar surgery knowledge graph based on historical surgical risk parameters, historical dental risk parameters, and historical health risk parameters. The embedding computation unit is used to perform feature embedding computation based on the nodes and edge weights of the alveolar surgery knowledge graph to construct the alveolar surgery knowledge graph. The alveolar surgery risk association matrix acquisition unit is used to obtain the alveolar surgery risk association matrix based on the alveolar surgery knowledge graph.

[0041] In this embodiment, historical surgical risk parameters are first obtained based on alveolar surgical procedure identifiers. Specifically, using alveolar surgical procedure identifiers as search criteria, all surgical cases with the same identifier are selected from the historical database. The probability of occurrence of risk events in these cases is statistically analyzed and used as historical surgical risk parameters. Here, risk events refer to the collective term for various intraoperative and postoperative complications and adverse events, such as tooth root fracture and postoperative infection. Historical surgical risk parameters can reflect the overall surgical risk level of different alveolar surgical procedure identifiers.

[0042] For example, if the alveolar surgery procedure is identified as D7210, and a total of 1000 surgical cases corresponding to this identification are retrieved from the historical database, among which 80 cases have risk events, then the historical surgical risk parameter = 80 / 1000 = 8%.

[0043] Secondly, due to differences in anatomical structure and location among different types of teeth, the specific risks faced during surgery vary. Therefore, it is necessary to obtain historical dental risk parameters based on tooth type. Specifically, using tooth type as the search criterion, all surgical cases matching the current tooth type are selected from the historical database. The probability of specific risk events occurring in all surgical cases for that type of tooth is then calculated as historical dental risk parameters. Here, risk events refer to the collective term for various surgical risk events related to tooth type, such as inferior alveolar nerve injury and intraoperative bleeding. Historical dental risk parameters can reflect the overall surgical risk level for different tooth types.

[0044] For example, if the tooth category is the mandibular first molar, and a total of 2,000 surgical cases of this type of tooth are retrieved from the historical database, of which 100 cases had risk events, then the historical tooth risk parameter = 100 / 2000 = 5%.

[0045] Secondly, the values ​​of oral health parameters directly reflect the health status of oral tissues, and surgical risks vary under different health conditions. Therefore, it is necessary to obtain historical health risk parameters based on oral health parameters. Specifically, using oral health parameters as a baseline value, a corresponding allowable error range is set, such as oral health parameters ± a certain threshold. This is used as a search criterion to filter all surgical cases that meet the search criteria from the historical database. The probability of occurrence of risk events in such cases is then statistically analyzed and used as historical health risk parameters. Here, risk events refer to the collective term for various surgical risk events related to oral health parameters, such as poor postoperative healing and wound infection. Historical health risk parameters can reflect the overall surgical risk level under different oral health parameters.

[0046] For example, if the oral health parameters include a bone density value of 256 and an alveolar bone height value of 12mm, and the allowable error range is set to 256±5 for bone density and 12±1mm for alveolar bone height, and 500 surgical cases that meet the allowable error range are retrieved from the historical database, among which 20 cases have risk events, then the historical health risk parameter = 20 / 500 = 4%.

[0047] Furthermore, based on historical surgical risk parameters, historical dental risk parameters, and historical health risk parameters, edge weights of the alveolar surgery knowledge graph are obtained. For example, a new comprehensive surgical risk node representing the overall risk level of risk events is added. Then, historical surgical risk parameters are used as edge weights between alveolar surgery procedure identifiers and comprehensive surgical risk nodes, historical dental risk parameters are used as edge weights between tooth categories and comprehensive surgical risk nodes, and historical health risk parameters are used as edge weights between oral health parameters and comprehensive surgical risk nodes. This allows the edge weights to directly quantify the comprehensive correlation between oral parameters, alveolar surgery procedure identifiers, and comprehensive surgical risks.

[0048] Furthermore, based on the nodes and edge weights of the alveolar surgery knowledge graph, feature embedding calculations are performed to construct the alveolar surgery knowledge graph. Specifically, the determined alveolar surgery knowledge graph nodes are first initially connected according to their actual relationships. Then, each node is input into a graph neural network (GNN) for feature embedding calculations, transforming the node's textual or numerical features into low-dimensional vectors. Then, combined with edge weights, the vectors of each node's neighboring nodes are weighted and aggregated to update the node's embedding vector. This allows the node vector to integrate its own features and the risk features of associated nodes, ultimately forming a structurally complete and feature-quantified alveolar surgery knowledge graph. The feature embedding calculation is the process of mapping the high-dimensional features of nodes to a low-dimensional vector space, aiming to preserve the core information of the nodes and simplify subsequent calculations.

[0049] Finally, based on the alveolar surgery knowledge graph, an alveolar surgery risk association matrix is ​​obtained. Specifically, a node embedding matrix is ​​constructed by arranging the updated embedding vectors of all nodes in a preset order; then, through matrix operations, the node embedding matrix is ​​multiplied by its transpose to obtain a square matrix, which serves as the alveolar surgery risk association matrix. The alveolar surgery risk association matrix can quantify the risk association strength between different nodes.

[0050] Furthermore, the alveolar surgery risk association matrix acquisition unit includes: The embedding vector generation unit is used to perform feature embedding calculations on nodes in the alveolar surgery knowledge graph using a graph neural network model, and generate an initial embedding vector for each node. The weighted aggregation unit is used to perform weighted aggregation on the embedding vectors of the neighboring nodes connected to each node based on the edge weights of the alveolar surgery knowledge graph, and update the embedding vector of each alveolar surgery knowledge graph node. The embedding matrix generation unit is used to combine the embedding vectors of the updated alveolar surgery knowledge graph nodes as row vectors to form a node embedding matrix. The alveolar surgery risk correlation matrix calculation unit is used to calculate the product of the node embedding matrix and its transpose matrix to obtain the alveolar surgery risk correlation matrix.

[0051] In this embodiment, a graph neural network (GNN) model is first used to perform feature embedding calculations on nodes in the alveolar surgery knowledge graph, generating an initial embedding vector for each node. Specifically, a GNN is a deep learning model specifically designed for processing graph-structured data. It effectively captures the relationships between nodes. Each node in the alveolar surgery knowledge graph (such as oral parameters, alveolar surgery procedure identifiers, and comprehensive surgical risk nodes) is input into a pre-trained GNN model. The model encodes and reduces the dimensionality of the original text or numerical features of each node, mapping the high-dimensional, heterogeneous original features to a low-dimensional vector space of uniform dimension, serving as the initial embedding vector for each node. The initial embedding vector has a preset fixed dimension, such as 64 or 128, and each element in the vector is a continuous numerical value. The initial embedding vector can retain the core feature information of the nodes while transforming discrete node data into a vector form that can be directly mathematically processed by the computer.

[0052] For example, the tooth category node "mandibular first molar" in the alveolar surgery knowledge graph is input into a graph neural network model (GNN). After encoding, the initial embedding vector generated is [0.123, 0.456, ..., 0.789]. The alveolar surgery procedure identifier node D7210 is encoded by the graph neural network model (GNN), and the initial embedding vector generated is [0.234, 0.567, ..., 0.890]. This achieves unified quantization of features of different types of nodes.

[0053] Secondly, based on the edge weights of the alveolar surgery knowledge graph, the embedding vectors of neighboring nodes connected to each node are weighted and aggregated to update the embedding vector of each alveolar surgery knowledge graph node. Neighboring nodes refer to other nodes in the knowledge graph that are directly connected to the current node through edges; these nodes have a direct risk association with the current node. The purpose of weighted aggregation is to ensure that each node's vector not only includes its own features but also reflects the risk impact of associated nodes, making the node features more comprehensive.

[0054] For example, by identifying all neighboring nodes of each node, the edge weights between the current node and each neighboring node are extracted. The edge weights are used as weighting coefficients to calculate the weighted neighboring node vector. Then, all the weighted neighboring node vectors are summed. Finally, the summation result is fused with the current node's own initial embedding vector, such as by weighted averaging, to obtain the updated embedding vector.

[0055] Next, the updated embedding vectors of the alveolar surgery knowledge graph nodes are used as row vectors and combined to form a node embedding matrix. For example, after weighted aggregation, the updated embedding vectors of all nodes have a uniform dimension, and each vector contains both the node's own features and neighborhood association features. To facilitate subsequent matrix operations, the updated embedding vectors of each node are arranged in a predetermined fixed order, such as tooth category nodes, oral health parameter nodes, alveolar surgery procedure identifier nodes, and comprehensive surgical risk nodes, forming a two-dimensional matrix as the node embedding matrix. The number of rows in the node embedding matrix equals the total number of nodes in the knowledge graph, and the number of columns equals the dimension of the embedding vectors. Each element in the node embedding matrix corresponds to the quantized value of a node on a certain feature dimension, fully preserving the feature information of all nodes and the association information between nodes.

[0056] For example, if the alveolar surgery knowledge graph contains five nodes: mandibular first molar, bone density value 256, alveolar bone height value 12mm, alveolar surgery procedure identifier D7210, and comprehensive surgical risk, and each updated embedding vector is 64-dimensional, then the node embedding matrix is ​​5 rows and 64 columns, with each row corresponding to the updated embedding vector of the above five nodes.

[0057] Finally, the product of the node embedding matrix and its transpose is calculated to obtain the alveolar surgery risk association matrix. Specifically, the node embedding matrix is ​​first transposed by swapping its rows and columns. For example, if the node embedding matrix is ​​M rows and N columns, its transpose is N rows and M columns. Then, following the matrix multiplication rules, the node embedding matrix and the transpose are multiplied to obtain an M-row, M-column square matrix, which serves as the alveolar surgery risk association matrix. The core function of the alveolar surgery risk association matrix is ​​to quantify the association strength between any two nodes in the alveolar surgery knowledge graph within the risk feature space.

[0058] In summary, compared to existing technologies, this application performs feature embedding calculations on the alveolar surgery atlas to construct an alveolar surgery knowledge graph and obtain an alveolar surgery risk association matrix. This achieves unified quantification and structured integration of heterogeneous node features in the alveolar surgery atlas, accurately characterizing the strength of deep risk associations between nodes, and providing a computable model foundation for subsequent risk assessment results based on individual user parameters.

[0059] The parameter acquisition module is used to collect the user's oral cavity parameters. Based on the oral cavity parameters, the risk correlation matrix of alveolar surgery is solved to obtain the initial risk assessment results of the surgery.

[0060] The alveolar surgery risk correlation matrix is ​​a holistic risk model built based on historical data, which cannot directly adapt to the differences in individual oral parameters of users. Moreover, traditional surgical risk assessment lacks precise quantitative application of individual oral characteristics. Therefore, it is necessary to collect user-specific oral parameters and combine them with the alveolar surgery risk correlation matrix to realize the transformation from a general risk model to a user-specific basic risk assessment.

[0061] To address the aforementioned issues, this application collects user oral parameters and, based on these parameters, solves the risk correlation matrix for alveolar surgery to obtain the initial risk assessment results for the surgery.

[0062] Specifically, the parameter acquisition module includes: The parameter acquisition unit is used to collect the user's oral cavity parameters and obtain historical oral cavity parameters based on the user's alveolar surgical procedure identifier. The deviation probability acquisition unit is used to acquire the deviation probability of alveolar surgery detection based on the user's oral parameters and historical oral parameters. The initial risk assessment unit is used to solve the alveolar surgery risk correlation matrix based on the user's oral parameters and the probability of deviation in alveolar surgery detection, and obtain the initial risk assessment results of the surgery.

[0063] In this embodiment, user oral parameters are first collected, and historical oral parameters are obtained based on the user's alveolar surgery procedure identifier. For example, the target user's oral parameters and alveolar surgery procedure identifier are collected from the patient's electronic medical record. Examples include a tooth type of mandibular first molar, oral health parameters of bone mineral density of 220 and alveolar bone height of 11mm, and an alveolar surgery procedure identifier of D7210. Then, based on the alveolar surgery procedure identifier D7210, historical oral parameters for all D7210 procedure cases are extracted from the historical database, ultimately forming a historical oral parameter set containing 5000 valid data points.

[0064] Secondly, alveolar surgery detection is obtained based on the user's oral parameters and historical oral parameters. For example, the mean and standard deviation of numerical parameters (such as bone density and alveolar bone height) in historical oral parameters can be calculated. Then, the probability of deviation in alveolar surgery detection for numerical parameters in the user's oral parameters can be calculated. Next, the historical frequency proportion of categorical parameters (such as tooth category) in historical oral parameters can be calculated. Subtracting this historical frequency proportion from 1 yields the probability of deviation in alveolar surgery detection for categorical parameters. Finally, the mean of the probability of deviation in alveolar surgery detection for numerical parameters and the probability of deviation in alveolar surgery detection for categorical parameters is calculated to obtain the overall probability of deviation in alveolar surgery detection.

[0065] For example, based on a historical oral parameter set containing 5000 valid data points, the mean of bone mineral density (BMD) was calculated to be 250 with a standard deviation of 30, and the mean of alveolar bone height was 12 mm with a standard deviation of 1.5 mm. The historical occurrence frequency of the mandibular first molar (tooth category) was calculated to be 60%. Therefore, the probability of deviation in alveolar surgical testing for BMD is |220-250| / 30 = 1.0, the probability of deviation for alveolar surgical testing for alveolar bone height is |11-12| / 1.5≈0.67, and the probability of deviation for alveolar surgical testing for tooth category is 1-60% = 0.4. Finally, the mean was calculated, resulting in the probability of deviation in alveolar surgical testing = (1.0+0.67+0.4) / 3≈0.69, thus achieving a unified quantification of the impact of deviations in the two types of parameters.

[0066] Finally, based on the user's oral cavity parameters and the probability of deviation in alveolar surgery detection, the risk correlation matrix for alveolar surgery was solved to obtain the initial risk assessment results. These initial risk assessment results focus on the fundamental risks arising from the user's individual oral cavity characteristics and do not include the inherent risk factors of the surgical procedure itself. They form the data basis for obtaining the final risk assessment results.

[0067] Furthermore, the initial risk assessment unit includes: The submatrix acquisition unit is used to acquire node-related submatrices in the alveolar surgery risk association matrix based on the user's oral parameters and the user's alveolar surgery procedure identifier. The calibration calculation unit is used to calibrate the node-related submatrix based on the probability of deviation in alveolar surgery detection, and obtain the risk-related submatrix. The typicality acquisition unit is used to obtain the occurrence rate of the risk association submatrix in the alveolar surgery risk association matrix, and to obtain the typicality of the surgical procedure. The oral cavity parameter vector construction unit is used to construct a user's oral cavity parameter vector based on the user's oral cavity parameters. The risk vector acquisition unit is used to acquire the user's risk vector based on the user's oral parameter vector and risk correlation submatrix; The risk assessment calculation unit is used to obtain the user risk assessment result based on the user risk vector, and to correct the user risk assessment result based on the typicality of the surgical procedure, so as to obtain the initial risk assessment result of the surgery.

[0068] In this embodiment, since the complete alveolar surgery risk association matrix contains risk association information of all nodes in the knowledge graph, the dimensions may reach hundreds of rows and columns, while the user only needs to focus on the nodes directly related to their own surgery, the node-related submatrix in the alveolar surgery risk association matrix is ​​obtained based on the user's oral parameters and the user's alveolar surgery procedure identifier.

[0069] Specifically, based on user oral parameters (such as mandibular first molar, bone density value of 220, and alveolar bone height value of 11mm) and user alveolar surgery procedure identifier (such as D7210), all rows and columns corresponding to these nodes are extracted from the complete alveolar surgery risk association matrix. The square matrix formed by the intersection of these rows and columns is the node-related submatrix. For example, a 4-row, 4-column node-related submatrix is ​​formed, which greatly simplifies the computation compared to the complete alveolar surgery risk association matrix.

[0070] Secondly, based on the probability of deviation in alveolar surgery detection, the node-related submatrix is ​​corrected to obtain the risk-related submatrix. Specifically, since a higher probability of deviation in alveolar surgery detection indicates a more significant difference between the user's oral parameters and historical oral parameters for the same procedure, and a higher uncertainty in surgical risk, each element value in the node-related submatrix is ​​multiplied by (1 + probability of deviation in alveolar surgery detection). The purpose of adding 1 is to avoid the element value remaining unchanged when the probability of deviation in alveolar surgery detection is 0 (i.e., the user's oral parameters are completely consistent with historical oral parameters), thus ensuring the integrity of the correction logic.

[0071] For example, if the value of an element in the node correlation submatrix is ​​0.05 and the probability of deviation in alveolar surgery detection is 0.69, then the corrected element value = 0.05 × (1 + 0.69) = 0.0845. This amplifies the element value in the node correlation submatrix, and the higher the probability of deviation in alveolar surgery detection, the higher the degree of correction. The corrected node correlation submatrix is ​​the risk correlation submatrix.

[0072] Next, the occurrence rate of the risk correlation submatrix within the alveolar surgery risk correlation matrix is ​​obtained to determine the typicality of the surgical procedure. For example, the sum of all elements in the risk correlation submatrix is ​​calculated, then the sum of all elements in the alveolar surgery risk correlation matrix is ​​calculated, and finally, the sum of all elements in the risk correlation submatrix is ​​divided by the sum of all elements in the alveolar surgery risk correlation matrix to obtain the typicality of the surgical procedure.

[0073] For example, if the sum of all elements in the risk association submatrix is ​​8.5 and the sum of all elements in the alveolar surgery risk association matrix is ​​85, then the typicality of the procedure is 8.5 / 85 = 0.1, indicating that the user's parameter combination occurs in only 10% of similar surgeries, which is a rare combination with high risk uncertainty.

[0074] Furthermore, a user oral cavity parameter vector is constructed based on the user's oral cavity parameters. Specifically, the user's oral cavity parameters are arranged in a fixed preset order, such as tooth category, bone density value, and alveolar bone height value, to form a user oral cavity parameter vector. The purpose of constructing the user oral cavity parameter vector is to transform the user oral cavity parameters, which contain categorical parameters (such as tooth category) and numerical parameters (such as bone density value and alveolar bone height value), into a vector of purely numerical form to meet the mathematical requirements of subsequent matrix operations.

[0075] Optionally, categorization parameters (such as tooth category) need to be converted into numerical values ​​through a preset unique coding rule: pre-assign exclusive integer codes to all possible tooth categories, for example, the maxillary central incisor is coded as 1, the third molar is coded as 2, ..., the mandibular first molar is coded as 5, etc.; numerical parameters (such as bone density value, alveolar bone height value) directly use the original measurement values ​​without additional conversion.

[0076] For example, if the user's oral parameters are the mandibular first molar, bone density value of 256, and alveolar bone height value of 12mm, the mandibular first molar is converted into the value 5 according to the preset unique coding rule. The bone density value of 256 and the alveolar bone height value of 12mm are directly adopted as the original measurement values ​​and arranged in a preset order to form the user's oral parameter vector: [5,220,11]. The user's oral parameter vector is in pure numerical form and can directly participate in subsequent matrix multiplication operations.

[0077] Furthermore, a user risk vector is obtained based on the user's oral cavity parameter vector and the risk correlation submatrix. Specifically, the user's oral cavity parameter vector is multiplied by the risk correlation submatrix to obtain the user risk vector. In this way, the individual characteristics of the user's oral cavity parameters are fused with the corrected risk correlation information, quantifying the contribution of each relevant node to the user's surgical risk. The risk correlation submatrix has a dimension of M rows and M columns, where M is the total number of user-related nodes; the user's oral cavity parameter vector has a dimension of 1 row and N columns, where N is the total number of user oral cavity parameters.

[0078] For example, before the calculation, it is necessary to ensure that the number of columns in the user's oral cavity parameter vector matches the number of rows in the risk association submatrix through vector expansion or matrix slicing. For instance, the dimension of the user's oral cavity parameter vector can be adjusted to 1 row and M columns by padding with zeros. During the calculation, the user's oral cavity parameter vector and the risk association submatrix are multiplied according to the matrix multiplication rules to obtain a one-dimensional vector of 1 row and M columns, which serves as the user's risk vector. Each element of the user's risk vector corresponds to the risk contribution value of a relevant node in the alveolar surgery knowledge graph. The larger the element value, the more significant the impact of the feature corresponding to that node (such as a specific tooth category or bone density value) on the user's surgical risk.

[0079] For example, if the user's oral parameter vector is [5,220,11], and the risk association submatrix is ​​5 rows and 5 columns, corresponding to 5 nodes: tooth category node, bone density node, alveolar bone height node, alveolar surgery procedure identifier node, and comprehensive surgical risk node, then the user's oral parameter vector needs to be expanded to 1 row and 5 columns. The expansion logic can be based on the complete node list of the risk association submatrix, supplementing the corresponding values ​​of the missing nodes in the user's oral parameter vector: the alveolar surgery procedure identifier node can use its predefined code, such as the code 3 corresponding to the alveolar surgery procedure identifier D7210, and the basic value of the comprehensive surgical risk node association uses a default constant, such as 1, to maintain the consistency of the calculation dimension. Therefore, the expanded user's oral parameter vector is [5,220,11,3,1]. Then, the expanded user's oral parameter vector is multiplied by the risk association submatrix according to the matrix multiplication rules to obtain the user's risk vector as [8.2,10.5,7.8,9.3,12.6]. The elements correspond to the risk contribution values ​​of the tooth category node, bone density node, alveolar bone height node, alveolar surgical procedure identification node, and comprehensive surgical risk node, respectively. The larger the element value, the more significant the impact of the node on the user's surgical risk.

[0080] Finally, user risk assessment results are obtained based on user risk vectors, and then corrected based on surgical procedure typicality to obtain the initial surgical risk assessment results. Specifically, user risk assessment results are obtained based on user risk vectors: the summation of all elements in the user risk vector is the user risk assessment result. The user risk assessment result integrates the contribution of each relevant node to the surgical risk and directly reflects the basic risk level brought about by the individual characteristics of the user's oral cavity. Surgical procedure typicality is an indicator reflecting the frequency of the combination of user oral cavity parameters and surgical procedure identifiers in historical cases, with a value of 0-1. The lower the frequency (i.e., the smaller the surgical procedure typicality), the less surgical experience there is for this combination, the higher the risk uncertainty, and the risk assessment result needs to be adjusted upward. Conversely, the higher the frequency (i.e., the greater the surgical procedure typicality), the lower the risk uncertainty, and the smaller the adjustment range.

[0081] For example, if the user's risk vector is [8.2, 10.5, 7.8, 9.3, 12.6], then the user risk assessment result = 8.2 + 10.5 + 7.8 + 9.3 + 12.6 = 48.4 points; the preset correction rule is: when the typicality of the procedure is ≤0.3, adjust by 15%; when the typicality of the procedure is ≤0.7, adjust by 5%; when the typicality of the procedure is >0.7, do not adjust.

[0082] For example, if a user's risk assessment score is 48.4 and the surgical procedure typicality is 0.1 (≤0.3), then it is corrected by 15%. The corrected initial surgical risk assessment score is 48.4 × (1 + 15%) = 55.66. The initial surgical risk assessment score reflects both the risk contribution of the user's individual characteristics and corrects for the uncertainty brought about by rare parameter combinations.

[0083] In summary, compared to existing technologies, this application collects user oral parameters and, based on these parameters, solves the alveolar surgical risk correlation matrix to obtain the initial surgical risk assessment results. This achieves a precise fit between a general alveolar surgical risk correlation matrix and the user's individual oral characteristics, quantifying the user's basic surgical risk based on their own oral parameters.

[0084] The risk assessment module is used to obtain surgical risk factors based on alveolar surgical procedure identifiers and, in combination with the initial surgical risk assessment results, to obtain the alveolar surgical risk assessment results.

[0085] The initial surgical risk assessment results obtained from the aforementioned steps only focus on the basic risks brought about by the individual oral characteristics of the user, and do not include the inherent risks of the alveolar surgery procedure itself, such as the complexity of the surgery. The differences in risk attributes of different procedures will directly affect the overall risk level of the surgery.

[0086] To address the aforementioned issues, this application obtains surgical risk factors based on alveolar surgical procedure identifiers and combines them with the initial surgical risk assessment results to obtain alveolar surgical risk assessment results.

[0087] Specifically, the risk assessment module includes: The surgical risk factor acquisition unit is used to acquire surgical risk factors based on alveolar surgical procedure identifiers, wherein the surgical risk factors include at least surgical complexity. The fusion assessment unit is used to obtain the risk assessment results of alveolar surgery based on surgical risk factors and combined with the initial surgical risk assessment results.

[0088] In this embodiment, surgical risk factors are first obtained based on alveolar surgical procedure identifiers. These surgical risk factors are quantitative indicators characterizing the inherent risk level of a specific alveolar surgical procedure, and include at least surgical complexity.

[0089] For example, the quantification logic of surgical complexity can be based on clinical treatment guidelines and historical surgical data: a unique complexity level code is pre-assigned to each alveolar surgical procedure identifier, such as levels 1-5, with higher levels indicating higher complexity. The level code is then converted into a corresponding quantification coefficient, such as level 1 corresponding to 1.0, level 2 to 1.2, level 3 to 1.5, level 4 to 1.8, and level 5 to 2.0. The formulation of the coding rules should refer to clinical guidelines. For example, simple loose tooth extraction (such as alveolar surgical procedure identifier D7110) has a complexity level of 1 and a quantification coefficient of 1.0 because it is simple to operate and has little trauma; impacted third molar extraction (such as alveolar surgical procedure identifier D7210) has a complexity level of 4 and a quantification coefficient of 1.8 because it requires cutting the gingiva and removing bone tissue, making it complex and high-risk.

[0090] In addition to surgical complexity, other surgical risk factors can be added according to the same acquisition logic to improve the accuracy and reliability of surgical risk assessment results.

[0091] Secondly, based on the surgical procedure risk factors and combined with the initial surgical risk assessment results, the alveolar surgical risk assessment results are obtained. For example, the surgical procedure risk factors can be multiplied by the initial surgical risk assessment results to obtain the alveolar surgical risk assessment results. Through multiplication calculation, the more complex the surgical procedure, the larger the surgical procedure risk factor, and the more significant the amplification effect on the initial surgical risk assessment results. If there are multiple surgical procedure risk factors, the weighted average of the multiple surgical procedure risk factors can be calculated first, and then multiplied by the initial surgical risk assessment results.

[0092] For example, if a user's initial surgical risk assessment score is 55.66, and the surgical procedure risk factor corresponding to the alveolar surgery procedure identifier D7210 is a surgical complexity of 1.8, then the final alveolar surgery risk assessment score = 55.66 × 1.8 ≈ 100.2. The alveolar surgery risk assessment result fully reflects both the basic risks arising from the user's individual oral characteristics (such as tooth type and bone density) and the inherent risks of the surgical procedure itself. A higher alveolar surgery risk assessment score indicates a higher overall risk level for the surgery: on the one hand, it means that the user's oral conditions (such as rare tooth types, abnormal bone density, insufficient alveolar bone height, etc.) lead to higher surgical difficulty and uncertainty; on the other hand, it indicates that the inherent risks of the selected surgical procedure (such as high surgical complexity, significant trauma, and high operational difficulty) are more significant. In this way, both individual differences and the inherent attributes of the surgical procedure are considered, providing clinicians with a comprehensive quantitative basis for developing surgical plans and mitigating risks.

[0093] In summary, compared to existing technologies, this application obtains surgical risk factors based on alveolar surgical procedure identification and combines them with the initial surgical risk assessment results to obtain alveolar surgical risk assessment results. This achieves a fusion of inherent surgical risks and individual user risks, overcoming the limitations of single assessments and yielding comprehensive and accurate alveolar surgical risk assessment results.

[0094] In summary, the embodiments of this application have at least the following technical effects: Compared to existing technologies, this application first obtains oral parameters and alveolar surgical procedure identifiers as nodes in the alveolar surgical knowledge graph. In this way, it integrates individual patient oral characteristics and surgical type characteristics that influence surgical risk, providing basic data units for the construction of the alveolar surgical knowledge graph.

[0095] Secondly, this application performs feature embedding calculations on the alveolar surgery atlas to construct an alveolar surgery knowledge graph and obtain an alveolar surgery risk association matrix. This achieves unified quantification and structured integration of heterogeneous node features in the alveolar surgery atlas, accurately characterizing the strength of deep risk associations between nodes, and providing a computable model foundation for subsequent risk assessment results based on individual user parameters.

[0096] Furthermore, this application collects user oral parameters and, based on these parameters, solves the alveolar surgical risk correlation matrix to obtain the initial surgical risk assessment results. In this way, a precise fit is achieved between the general alveolar surgical risk correlation matrix and the user's individual oral characteristics, quantifying the user's basic surgical risk based on their own oral parameters.

[0097] Finally, this application obtains surgical risk factors based on alveolar surgical procedure identifiers and combines them with the initial surgical risk assessment results to obtain alveolar surgical risk assessment results. In this way, it integrates the inherent risks of the surgical procedure with the individual user's risk profile, overcoming the limitations of a single assessment and obtaining comprehensive and accurate alveolar surgical risk assessment results.

[0098] Through the aforementioned technical solution, this application transforms oral parameters and alveolar surgical procedure identifiers into knowledge graph nodes. A risk association matrix for alveolar surgery is constructed through feature embedding calculations. The initial surgical risk assessment results are then calculated by combining individual user oral characteristics, and surgical risk factors are incorporated to complete the final assessment. This effectively overcomes the shortcomings of traditional assessments, which rely on experience, are highly subjective, and lack standardized criteria. It achieves a systematic, objective, and comprehensive quantification of alveolar surgical risks, accurately links individual differences with inherent surgical risks, improves the accuracy of risk prediction, and provides reliable support for surgical plan optimization and surgical safety assurance, aligning with the needs of precision medicine development.

[0099] To achieve the above-mentioned objectives, the present invention also provides an electronic device comprising: a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the above-mentioned method for assessing the risk of alveolar surgery.

[0100] To achieve the above-mentioned objectives, the present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-mentioned method for assessing the risk of alveolar surgery.

[0101] Those skilled in the art will recognize that the modules and algorithm steps described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0102] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the above-described apparatus and equipment can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0103] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0104] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of the embodiments of the present invention, depending on actual needs.

[0105] In addition, the functional modules in the embodiments of the present invention can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0106] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion 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 energy-saving signal transmission / reception 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, ROM, RAM, magnetic disks, or optical disks.

[0107] The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the inventive concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this application.

[0108] It should be understood that the sequence number of each step in the invention and embodiments of the present invention does not absolutely imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

Claims

1. A method for assessing the risk of alveolar surgery, characterized in that, include: Obtain oral parameters and alveolar surgical procedure identifiers, and use these parameters and identifiers as nodes in the alveolar surgical knowledge graph. Historical surgical risk parameters, historical tooth risk parameters, and historical health risk parameters are obtained based on alveolar surgical procedure identifiers, tooth categories, and oral health parameters, and are used as edge weights for the alveolar surgical knowledge graph. Feature embedding calculations are performed based on nodes and edge weights of the alveolar surgery knowledge graph to construct the alveolar surgery knowledge graph and obtain the alveolar surgery risk association matrix. Collect user oral parameters and obtain historical oral parameters corresponding to alveolar surgery procedure identifiers. Based on user oral parameters and historical oral parameters, obtain the probability of alveolar surgery detection deviation. The risk correlation matrix of alveolar surgery is solved based on user oral parameters and the probability of deviation in alveolar surgery detection to obtain the initial risk assessment results of the surgery. Based on the surgical procedure identifier, surgical procedure risk factors are obtained, and the surgical procedure risk factors are integrated with the initial surgical risk assessment results to obtain the surgical risk assessment results of alveolar surgery.

2. The method for assessing the risk of alveolar surgery as described in claim 1, characterized in that, Oral parameters should include at least tooth category and oral health parameters; Based on oral parameters and alveolar surgical procedure identifiers as nodes in the alveolar surgical knowledge graph, the tooth category, bone mineral density value, alveolar bone height value and alveolar surgical procedure identifier in oral health parameters are each treated as an independent node; Tooth categories are extracted from electronic medical records or imaging reports, bone mineral density values ​​are calculated using the average grayscale value of CBCT image pixels, and alveolar bone height values ​​are calculated using the vertical distance between two points on the CBCT image.

3. The method for assessing the risk of alveolar surgery as described in claim 2, characterized in that, Based on the identification of alveolar surgical procedures, the probability of risk events occurring in historical cases of the same surgical procedure is used as a historical surgical risk parameter. Historical dental risk parameters are based on the probability of occurrence of risk events corresponding to different tooth categories. The probability of occurrence of risk events is statistically analyzed based on oral health parameters and used as historical health risk parameters; The edge weights between nodes in the alveolar surgery knowledge graph are calculated based on historical surgical risk parameters, historical dental risk parameters, and historical health risk parameters, thereby representing the probability of risk association between nodes.

4. The method for assessing the risk of alveolar surgery as described in claim 3, characterized in that, A graph neural network model is used to perform feature embedding calculations on nodes in the alveolar surgery knowledge graph to generate initial embedding vectors for each node. The embedding vector of the target node is updated by weighting and aggregating the embedding vectors of neighboring nodes connected to the target node based on edge weights. The updated node embedding vectors are combined as row vectors to form a node embedding matrix. The product of the node embedding matrix and its transpose is calculated to obtain the alveolar surgery risk association matrix, which represents the association strength of any node pair in the risk feature space.

5. The method for assessing the risk of alveolar surgery as described in claim 4, characterized in that, Obtain historical oral parameter sets for the same alveolar surgical procedure based on the alveolar surgical procedure identifier; The mean and standard deviation of numerical parameters in the historical oral parameter set are calculated, and the probability of numerical deviation is determined based on the degree of deviation between the user's numerical oral parameters and the mean. The historical frequency percentage of the classification parameters is statistically analyzed, and the probability of classification deviation is determined by the complementary value of the percentage. The numerical deviation probability and the categorical deviation probability are summarized to obtain the alveolar surgical detection deviation probability, which is used to represent the degree of difference between the user's oral parameters and the historical oral parameters of the same procedure.

6. The method for assessing the risk of alveolar surgery as described in claim 5, characterized in that, Based on user oral parameters and alveolar surgical procedure identifiers, node-related submatrices of corresponding nodes are extracted from the alveolar surgical risk association matrix; The risk association submatrix is ​​obtained by correcting the node-related submatrix based on the probability of deviation in alveolar surgery detection; The typicality of a surgical procedure is obtained by obtaining the probability of occurrence of the risk correlation submatrix within the risk correlation matrix of alveolar surgery. A user oral cavity parameter vector is constructed based on the user's oral cavity parameters, and a matrix operation is performed between the user oral cavity parameter vector and the risk correlation submatrix to obtain the user risk vector. The user risk assessment results are obtained based on the user risk vector, and then the user risk assessment results are corrected based on the typicality of the surgical procedure to obtain the initial surgical risk assessment results.

7. The method for assessing the risk of alveolar surgery as described in claim 6, characterized in that, The surgical procedure risk factor includes at least the numerical coefficient corresponding to the surgical complexity, and a comprehensive surgical procedure risk factor is obtained based on the numerical coefficient corresponding to the surgical complexity and other surgical procedure risk factors. When there are multiple surgical procedure risk factors, a weighted average of the multiple surgical procedure risk factors is taken to obtain a comprehensive surgical procedure risk factor. The risk assessment results of alveolar surgery are obtained by integrating the comprehensive surgical procedure risk factors with the initial surgical risk assessment results.

8. A risk assessment system for alveolar surgery, characterized in that, include: Alveolar surgery knowledge graph edge node construction module: Obtain oral parameters and alveolar surgery procedure identifiers, and use oral parameters and alveolar surgery procedure identifiers as nodes of alveolar surgery knowledge graph; Alveolar surgery knowledge graph edge weight construction module: Based on alveolar surgery procedure identifier, tooth category and oral health parameters, historical surgical risk parameters, historical tooth risk parameters and historical health risk parameters are obtained respectively, and used as edge weights of alveolar surgery knowledge graph; Alveolar surgery risk association matrix acquisition module: Based on the nodes and edge weights of the alveolar surgery knowledge graph, feature embedding calculation is performed to construct the alveolar surgery knowledge graph and obtain the alveolar surgery risk association matrix; Alveolar surgery detection deviation probability generation module: Collects user oral parameters and obtains historical oral parameters corresponding to alveolar surgery procedure identifiers, and generates the alveolar surgery detection deviation probability based on user oral parameters and historical oral parameters; The module for obtaining the initial surgical risk assessment results solves the correlation matrix of alveolar surgery risk based on the user's oral parameters and the probability of deviation in alveolar surgery detection, thus obtaining the initial surgical risk assessment results. Alveolar surgery risk assessment result acquisition module: Based on the alveolar surgery procedure identifier, the procedure risk factors are obtained, and the procedure risk factors are integrated with the initial surgical risk assessment results to obtain the alveolar surgery risk assessment results.

9. An electronic device, characterized in that, It includes a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements a method for assessing the risk of alveolar surgery as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, which, when executed by a processor, implements a method for assessing the risk of alveolar surgery as described in any one of claims 1 to 7.

Images