Endometrial cancer metastasis risk prediction method, system and equipment based on synergistic effect
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- THE FIRST AFFILIATED HOSPITAL ZHEJIANG UNIV COLLEGE OF MEDICINE
- Filing Date
- 2026-02-06
- Publication Date
- 2026-05-26
Smart Images
Figure CN122091198A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of disease prediction, and in particular to a method, system, and device for predicting the risk of metastasis of endometrial cancer based on synergistic effects. Background Technology
[0002] Current disease prediction models are primarily built upon the main effects of variables, neglecting the complex and substantial synergistic effects among clinical variables. These synergistic effects significantly impact model construction and interpretation. Detecting synergistic relationships between variables is highly beneficial for model construction, interpretability, and domain knowledge extraction. In endometrial cancer, the latest clinical guidelines emphasize molecular subtyping and the synergistic relationships between molecules. For example, the "double whammy" of PTEN (tumor suppressor gene inactivation) and PIK3CA (oncogene activation) directly targets cancer development. The synergy between POLE mutations and normal MMR function ensures a high mutational burden while preventing excessive genomic disturbances that lead to tumor cell death, thus maintaining tumor survival. Conversely, the synergistic amplification of TP53 and HER2 / CCNE1 in p53 abnormalities suggests a worse prognosis, requiring enhanced postoperative adjuvant therapy.
[0003] In model building, identifying synergistic features is a crucial part of feature engineering. Currently, there are several mainstream methods for identifying synergistic relationships: 1) Direct methods: These do not rely on complex models and directly detect synergistic relationships through data partitioning (e.g., the FAST index) or statistical tests (e.g., two-way ANOVA), offering a simple logic. 2) Model-specific methods: These methods automatically build a specific model from a given dataset and extract interaction information. They construct fixed models (e.g., GROVE, GUIDE) and compare the model performance of target feature pairs to deduce synergistic effects, but they are computationally expensive and have poor adaptability. 3) Non-model-specific methods: These methods do not restrict model types and determine synergistic relationships through error analysis or partial dependency graphs. The former depends on model time consumption, while the latter requires variable independence and makes it difficult to quantify the strength of synergy.
[0004] In summary, the above methods have two major drawbacks: first, they cannot distinguish between synergistic effects and main effects, and are prone to overestimating variable pairs with strong main effects; second, they are highly dependent on the model, computationally time-consuming, unsuitable for large datasets, and difficult to implement efficiently and accurately. Summary of the Invention
[0005] The purpose of this application is to provide a method, system, and device for predicting the risk of endometrial cancer metastasis based on synergistic effects, which can effectively improve the accuracy of risk prediction.
[0006] To achieve the above objectives, this application provides the following solution: Firstly, this application provides a method for predicting the risk of endometrial cancer metastasis based on synergistic effects, including: Obtain user case sample datasets; each user case sample dataset includes the user's medical characteristics and lymph node metastasis status; Based on the user case sample dataset, main effects were extracted to identify multiple key clinical features; Based on the user case sample dataset and multiple key clinical features, synergistic effects are extracted, and multiple new synergistic features are constructed. Based on multiple key clinical features and multiple synergistic new features, the user case sample dataset is modified by adding or subtracting data to obtain a new sample dataset. Then, based on the new sample dataset, a prediction model for the risk of endometrial cancer metastasis is constructed using random forest. When new user case sample data is collected, multiple key clinical features and multiple synergistic new features are extracted and then input into the endometrial cancer metastasis risk prediction model to obtain prediction results.
[0007] Secondly, this application provides a synergistic effect-based endometrial cancer metastasis risk prediction system, comprising: The sample data acquisition module is used to acquire user case sample datasets; each user case sample dataset includes the user's medical characteristics and lymph node metastasis status. The main effect extraction module is used to extract main effects based on the user case sample dataset to determine multiple key clinical features; The synergy effect extraction module is used to extract synergy effects and construct multiple new synergy features based on the user case sample dataset and multiple key clinical features. The risk prediction model construction module is used to obtain a new sample dataset by adding or subtracting data from the user case sample dataset based on multiple key clinical features and multiple synergistic new features, and then construct an endometrial cancer metastasis risk prediction model based on the new sample dataset using random forest. The risk prediction model uses a module to extract multiple key clinical features and multiple synergistic new features when new user case sample data is collected, and then inputs them into the endometrial cancer metastasis risk prediction model to obtain prediction results.
[0008] Thirdly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement a method for predicting the risk of endometrial cancer metastasis based on synergistic effects.
[0009] According to the specific embodiments provided in this application, the following technical effects are disclosed: This application extracts main effects based on user case sample datasets to determine multiple key clinical features; it extracts synergistic effects based on user case sample datasets and multiple key clinical features to construct multiple new synergistic features. Through the above processing, this application can distinguish between synergistic effects and main effects, and fully utilize the synergistic effects between variables. On this basis, based on the corresponding new sample dataset, a prediction model for endometrial cancer metastasis risk is constructed using random forest, and the prediction is performed using this model, which can effectively improve the accuracy of the model in predicting the risk of endometrial cancer metastasis. Attached Figure Description
[0010] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0011] Figure 1 This is an application environment diagram of the endometrial cancer metastasis risk prediction method based on synergistic effect in one embodiment of this application.
[0012] Figure 2 This is a flowchart illustrating a method for predicting the risk of endometrial cancer metastasis based on synergistic effects in one embodiment of this application.
[0013] Figure 3 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation
[0014] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.
[0015] This application proposes a direct detection method for relative synergistic effects based on information entropy. It addresses the problem of main effect interference by defining an index as the ratio of "information gain to the information content of the variable itself." The method employs data discretization to calculate information gain, overcoming the challenges of calculating continuous variables, and has been verified to possess universality, fairness, and robustness. Furthermore, it incorporates pairs of variables exhibiting synergistic effects as new feature variables into the model construction, effectively improving the model's accuracy in predicting the risk of endometrial cancer metastasis.
[0016] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0017] The synergistic effect-based method for predicting the metastatic risk of endometrial cancer provided in this application can be applied to, for example... Figure 1 In the application environment shown, terminal 101 communicates with server 102 via a network. A data storage system can store the data that server 102 needs to process. The data storage system can be set up independently, integrated into server 102, or placed in the cloud or on another server. Terminal 101 can send user case sample datasets or new user case sample data to server 102. After receiving the data, server 102 extracts the main effects and synergistic effects based on the user case sample dataset. Then, based on multiple key clinical features and multiple new synergistic features, it adds or subtracts data from the user case sample dataset to obtain a new sample dataset. A random forest is then used to construct an endometrial cancer metastasis risk prediction model. For new user case sample data, multiple key clinical features and multiple new synergistic features are extracted and then input into the endometrial cancer metastasis risk prediction model to obtain the prediction result. Server 102 can provide feedback on the endometrial cancer metastasis risk prediction model or the prediction result to terminal 101. Furthermore, in some embodiments, the endometrial cancer metastasis risk prediction method based on synergistic effects can also be implemented independently by server 102 or terminal 101.
[0018] The terminal 101 can be, but is not limited to, various desktop computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, and smart in-vehicle devices. Portable wearable devices can include smartwatches, smart bracelets, and head-mounted devices. The server 102 can be implemented using a standalone server or a server cluster composed of multiple servers, or it can be a cloud server.
[0019] In one exemplary embodiment, such as Figure 2 As shown, a method for predicting the risk of metastasis of endometrial cancer based on synergistic effects is provided. This method is executed by computer equipment, which can be executed by a computer device such as a terminal or a server alone, or by a terminal and a server together.
[0020] Clinical big data possesses complex characteristics with multiple features. Besides individual variables, the synergistic effects between variables can also significantly impact clinical outcomes, and the relationships between variables may provide solutions to clinical problems. Detecting synergistic relationships between variables is helpful for constructing medical disease models, improving model interpretability, and extracting domain knowledge. In this embodiment, the method is applied to... Figure 1Taking server 102 as an example, the explanation includes the following steps 201 to 205.
[0021] Step 201: Obtain the user case sample dataset; each user case sample dataset includes the user's medical characteristics and lymph node metastasis status. In specific applications, step 201 includes the following processing steps: (11) Collect multiple original user case sample data, namely clinical data of endometrial cancer.
[0022] (12) Perform missing value cleaning, normalization and encoding of classification variables on all the original user case sample data to obtain structured user medical features; any user medical feature and the corresponding lymph node metastasis status constitute a user case sample data; multiple user case sample data constitute a user case sample dataset.
[0023] The missing values can be cleaned using multiple imputation, and normalization is performed using Z-score normalization. Categorical variables can be encoded using one-hot encoding. The final output is a user case sample dataset X={X1,X2,...,XN}, where X1 is the medical data of the first patient (including 210 features such as age, BMI, and gene expression), X2 is the medical data of the second patient, and XN is the medical data of the Nth patient. The label set Y represents the lymph node metastasis status Y={0,1}, which is used to characterize the risk of endometrial cancer metastasis.
[0024] Step 202: Based on the user case sample dataset, perform main effect extraction to determine multiple key clinical features.
[0025] The main effect of each feature, i.e., the magnitude of the influence of the independent variables on the dependent variable. Key variables can be quickly located by identifying features with significant main effects, allowing for a detailed analysis of the intrinsic relationship between the independent and dependent variables and the specific manifestation of that relationship. In a specific application, step 202 includes: (21) Based on the user case sample dataset, a decision tree model is constructed using a gradient descent tree model.
[0026] (22) Based on the decision tree model, the contribution of each feature in the user's medical features to the lymph node metastasis status is calculated using the Gini coefficient.
[0027] For any node in the decision tree model, the following formula is used to calculate... coefficient: .
[0028] Where K is the number of feature categories in a node. denoted as the proportion of the k-th type of feature in the node.
[0029] The gain after eigenvalue splitting is calculated using the following formula. Used to measure the degree of improvement in node purity: .
[0030] in, The Gini coefficient of the parent node before the split; The number of samples in the i-th child node after splitting; The total number of samples in the parent node before the split; Let be the Gini coefficient of the i-th child node after splitting.
[0031] The contribution of each feature F to the lymph node metastasis status is calculated using the following formula. Specifically, it is the weighted sum of the gains of feature F across all split nodes: .
[0032] (23) Based on the contribution of each feature in the user's medical characteristics to the lymph node metastasis status, multiple key clinical features are determined and used as main fact features to find synergistic effect feature pairs and to construct the final model; among them, there are 20 key clinical features, including age, menopause status, BMI, G grade, pathological type and CLIC1, TESC, MLH1, PMS2, CDK1, BIRC7, IGF2, CYP51A1, DLK1, SSC4D, LMBR1, RAET1G, PTEN, PIK3CA, PLAC1 genes.
[0033] Step 203: Based on the user case sample dataset and multiple key clinical features, extract synergistic effects and construct multiple new synergistic features.
[0034] In one specific application, step 203 includes: (31) Based on the user’s medical characteristics and multiple clinical key characteristics, multiple feature pairs are constructed; wherein each feature pair includes a clinical key characteristic and a non-clinical key characteristic.
[0035] (32) For any pair of features, calculate the feature information gain and calculate the synergistic effect based on the feature information gain.
[0036] In short, for the extracted main effects, the feature information gain for predicting lymph node metastasis in each main effect feature pair without a main effect is calculated. The continuous variable is discretized in multiple dimensions (e.g., equal-frequency binning), and the formula for calculating the feature information gain of the feature pair (X1, X2) is as follows: G(Y|X2 to X1)=H(Y||X1,X2)-H(Y||X1).
[0037] Wherein, X1 is the key clinical feature, and X2 is the key non-clinical feature; Y is the outcome variable, referring to the lymph node metastasis status; H(Y||X1) is the main effect of X1 in predicting Y; H(Y||X1,X2) is the total effect of X1 and X2 in predicting Y; and G(Y|X2toX1) is the feature information gain provided by X2 to X1.
[0038] The above calculations can output information gain values for 3800 pairs of features, which can be used for synergistic effect screening.
[0039] (33) Based on the synergy effect, multiple effective synergy effect variable pairs are selected from all the feature pairs. The synergy effect in feature engineering refers to the "new predictive information" generated after combining multiple features. This application calculates the synergy effect (SE) using the following method: SE=MAX ( ).
[0040] Where G1=G(Y|X1 to X2), is the feature information gain provided by X1 to X2; G2=G(Y|X2 to X1), is the feature information gain provided by X2 to X1; M1=H(Y||X1), is the main effect of X1 in predicting Y; M2=H(Y||X2), is the main effect of X2 in predicting Y.
[0041] Based on the above calculations, a total of 8 pairs of effective synergistic variable pairs were selected. Several pairs of effective synergistic variable pairs include: age-parity, BMI-pathological type, menopause status-parity, menopause status-weight, MLH1-MGARP, CDK1-NEU1, IGF2-PGRMC1, and PTEN-weight.
[0042] (34) Based on any of the effective synergistic effect variable pairs, extract the synergistic effect as a new synergistic feature using a decision tree model. Specifically, input the effective synergistic effect variable pairs obtained in the previous step into the decision tree model to generate a new feature f(Xi,Xj): Fs=∑f(Xi,Xj); where f is the synergistic prediction value output by the decision tree (such as the joint risk score of age-parity), which is also the new synergistic feature.
[0043] Step 204: Based on multiple key clinical features and multiple synergistic new features, the user case sample dataset is modified by adding or subtracting data to obtain a new sample dataset. Then, based on the new sample dataset, a prediction model for the risk of endometrial cancer metastasis is constructed using random forest.
[0044] Based on the new sample dataset, a model for predicting the risk of endometrial cancer metastasis is constructed using random forest, including: (41) Construct a prediction model using random forest: Model = RandomForest(Fm∪Fs); where Fm is the key clinical feature and Fs is the new collaborative feature; Model refers to the prediction model and RandomForest refers to the random forest.
[0045] (42) The new sample dataset is divided into a training set, a validation set, and a test set in a 7:2:1 ratio. The prediction model is trained using the training set, its parameters are tuned using the validation set, and it is evaluated using the test set to obtain the final endometrial cancer metastasis risk prediction model. In this process, a gradient descent decision tree is selected as the base model for training and fitting.
[0046] In addition, during the validation process, the stability and predictive performance of the model were verified through 5-fold cross-validation and compared with the predictive model established solely through main effects, thereby further verifying the effectiveness and accuracy of the model in this application.
[0047] Step 205: When new user case sample data is collected, multiple corresponding key clinical features and multiple synergistic new features are extracted and then input into the endometrial cancer metastasis risk prediction model to obtain the prediction results.
[0048] Currently, most data-based knowledge extraction focuses on the main effect of each feature. However, in addition to the main effect, synergistic effects are also a crucial aspect of knowledge extraction. Since synergistic effects are often masked by the main effect, they need to be mined. This application focuses on the synergistic effect between clinical variables, distinguishing and identifying the synergistic effect between clinical features and the main effect. Based on the identification of synergistic effects, during model construction, variable pairs with synergistic effects are found, and the synergistic effect is extracted as a new synergistic feature using a decision tree model for subsequent disease model building, effectively improving model accuracy. In short, this application can detect synergistic relationships between clinical data and, in the clinical disease of endometrial cancer, establish a lymph node metastasis prediction model based on the discovered synergistic relationships, improving the model's prediction accuracy.
[0049] Based on the same inventive concept, this application also provides a system. The solution provided by this system is similar to the solution described in the above method. Therefore, the specific limitations of one or more system embodiments provided below can be found in the limitations of the method above, and will not be repeated here.
[0050] In one exemplary embodiment, a synergistic effect-based endometrial cancer metastasis risk prediction system is provided, comprising: The sample data acquisition module is used to acquire user case sample datasets; each user case sample dataset includes the user's medical characteristics and lymph node metastasis status.
[0051] The main effect extraction module is used to extract main effects based on the user case sample dataset to determine multiple key clinical features.
[0052] The synergy effect extraction module is used to extract synergy effects and construct multiple new synergy features based on the user case sample dataset and multiple key clinical features.
[0053] The risk prediction model construction module is used to obtain a new sample dataset by adding or subtracting data from the user case sample dataset based on multiple key clinical features and multiple synergistic new features, and then construct an endometrial cancer metastasis risk prediction model based on the new sample dataset using random forest.
[0054] The risk prediction model uses a module to extract multiple key clinical features and multiple synergistic new features when new user case sample data is collected, and then inputs them into the endometrial cancer metastasis risk prediction model to obtain prediction results.
[0055] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 3 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and databases. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a synergistic effect-based method for predicting the risk of endometrial cancer metastasis.
[0056] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.
[0057] In one exemplary embodiment, a computer device is also provided, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps in the above-described method embodiments.
[0058] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.
[0059] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.
[0060] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.
[0061] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).
[0062] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.
[0063] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0064] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.
Claims
1. A method for predicting the risk of metastasis in endometrial cancer based on synergistic effects, characterized in that, The method includes: Obtain user case sample datasets; each user case sample dataset includes the user's medical characteristics and lymph node metastasis status; Based on the user case sample dataset, main effects were extracted to identify multiple key clinical features; Based on the user case sample dataset and multiple key clinical features, synergistic effects are extracted, and multiple new synergistic features are constructed. Based on multiple key clinical features and multiple synergistic new features, the user case sample dataset is modified by adding or subtracting data to obtain a new sample dataset. Then, based on the new sample dataset, a prediction model for the risk of endometrial cancer metastasis is constructed using random forest. When new user case sample data is collected, multiple key clinical features and multiple synergistic new features are extracted and then input into the endometrial cancer metastasis risk prediction model to obtain prediction results.
2. The method for predicting the risk of endometrial cancer metastasis based on synergistic effects according to claim 1, characterized in that, Based on the aforementioned user case sample dataset, main effects were extracted to identify several key clinical features, including: Based on the aforementioned user case sample dataset, a decision tree model is constructed using a gradient descent tree model. Based on the decision tree model, the contribution of each feature in the user's medical characteristics to the lymph node metastasis status is calculated using the Gini coefficient. Based on the contribution of each of the user's medical characteristics to the lymph node metastasis status, multiple key clinical characteristics are determined; among them, multiple key clinical characteristics include age, menopause status, BMI, G grade, pathological type, and CLIC1, TESC, MLH1, PMS2, CDK1, BIRC7, IGF2, CYP51A1, DLK1, SSC4D, LMBR1, RAET1G, PTEN, PIK3CA, and PLAC1 genes.
3. The method for predicting the risk of endometrial cancer metastasis based on synergistic effects according to claim 2, characterized in that, Based on the decision tree model, the contribution of each feature in the user's medical characteristics to the lymph node metastasis status is calculated using the Gini coefficient, including: For any node in the decision tree model, the following formula is used to calculate... coefficient: ； Where K is the number of feature categories in a node. The proportion of the k-th feature in the node; The gain after eigenvalue splitting is calculated using the following formula. : ； in, The Gini coefficient of the parent node before the split; The number of samples in the i-th child node after splitting; The total number of samples in the parent node before the split; Let Gini coefficient be the value of the i-th child node after splitting. The contribution of each feature F to the lymph node metastasis status is calculated using the following formula. : 。 4. The method for predicting the risk of endometrial cancer metastasis based on synergistic effects according to claim 1, characterized in that, Based on the user case sample dataset and multiple key clinical features, synergistic effects are extracted, and several new synergistic features are constructed, including: Based on the user's medical characteristics and multiple key clinical characteristics, multiple feature pairs are constructed; wherein each feature pair includes one key clinical characteristic and one key non-clinical characteristic. For any pair of features, calculate the feature information gain, and calculate the synergistic effect based on the feature information gain; Based on the synergistic effect, multiple effective synergistic effect variable pairs are selected from all the feature pairs; Based on any of the aforementioned effective synergistic effect variable pairs, the synergistic effect is extracted as a new synergistic feature using a decision tree model.
5. The method for predicting the risk of endometrial cancer metastasis based on synergistic effects according to claim 4, characterized in that, The formula for calculating the feature information gain is as follows: G(Y|X2 to X1)=H(Y||X1,X2)-H(Y||X1); Wherein, X1 is the key clinical feature, X2 is the key non-clinical feature; Y is the outcome variable, referring to the lymph node metastasis status; H(Y||X1) is the main effect of X1 in predicting Y; H(Y||X1,X2) is the total effect of X1 and X2 in predicting Y; G(Y|X2 toX1) is the feature information gain provided by X2 to X1. The formula for calculating the synergistic effect SE is as follows: SE=MAX( ); Where G1=G(Y|X1 to X2), is the feature information gain provided by X1 to X2; G2=G(Y|X2 to X1), is the feature information gain provided by X2 to X1; M1=H(Y||X1), is the main effect of X1 in predicting Y; M2=H(Y||X2), is the main effect of X2 in predicting Y.
6. The method for predicting the risk of endometrial cancer metastasis based on synergistic effect according to claim 4, characterized in that, Several effective synergistic effect variable pairs include: age-parity, BMI-pathological type, menopausal status-parity, menopausal status-weight, MLH1-MGARP, CDK1-NEU1, IGF2-PGRMC1, and PTEN-weight.
7. The method for predicting the risk of endometrial cancer metastasis based on synergistic effects according to claim 1, characterized in that, Based on the new sample dataset, a model for predicting the risk of endometrial cancer metastasis was constructed using random forest, including: A prediction model is constructed using random forest: Model = RandomForest(Fm∪Fs); where Fm represents key clinical features and Fs represents novel collaborative features; The new sample dataset is divided into a training set, a validation set, and a test set. The prediction model is trained using the training set, tuned using the validation set, and evaluated using the test set to obtain the final endometrial cancer metastasis risk prediction model.
8. The method for predicting the risk of endometrial cancer metastasis based on synergistic effect according to claim 1, characterized in that, Obtain the user medical record sample dataset, including: Collect data from multiple original user case samples; All the original user case sample data are cleaned of missing values, normalized, and categorically encoded to obtain structured user medical features; any user medical feature and its corresponding lymph node metastasis status constitute a user case sample data; multiple user case sample data constitute a user case sample dataset.
9. A synergistic effect-based system for predicting the risk of endometrial cancer metastasis, characterized in that, The system includes: The sample data acquisition module is used to acquire user case sample datasets; each user case sample dataset includes the user's medical characteristics and lymph node metastasis status. The main effect extraction module is used to extract main effects based on the user case sample dataset to determine multiple key clinical features; The synergy effect extraction module is used to extract synergy effects and construct multiple new synergy features based on the user case sample dataset and multiple key clinical features. The risk prediction model construction module is used to obtain a new sample dataset by adding or subtracting data from the user case sample dataset based on multiple key clinical features and multiple synergistic new features, and then construct an endometrial cancer metastasis risk prediction model based on the new sample dataset using random forest. The risk prediction model uses a module to extract multiple key clinical features and multiple synergistic new features when new user case sample data is collected, and then inputs them into the endometrial cancer metastasis risk prediction model to obtain prediction results.
10. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the synergistic effect-based method for predicting the risk of endometrial cancer metastasis as described in any one of claims 1-8.