Ovarian cancer risk assessment auxiliary system based on artificial intelligence

Abstract

The invention discloses an ovarian cancer risk assessment auxiliary system based on artificial intelligence. The ovarian cancer risk assessment auxiliary system comprises a data acquisition module, an ovarian feature extraction module, an ovarian cancer risk assessment model construction module, a model performance improvement module and an ovarian cancer risk assessment auxiliary module. The invention relates to the technical field of cancer risk assessment, in particular to an ovarian cancer risk assessment auxiliary system based on artificial intelligence, which improves the accuracy and comprehensiveness of feature extraction through detail enhancement, contour extraction, morphological feature calculation and textural feature calculation; by constructing a double-layer fusion kernel function, designing an innovative penalty term, generating a target function and determining an optimal condition, the classification performance and stability of the model are improved, and the accuracy and reliability of ovarian cancer risk assessment are improved; by constructing a diversity maintenance factor, designing a parameter competition mechanism, compensating a step length and performing crossover variation, the efficiency and stability of model performance improvement are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of cancer risk assessment, and specifically refers to an artificial intelligence-based ovarian cancer risk assessment assistance system. Background Art

[0002] An artificial intelligence-based ovarian cancer risk assessment assistance system is an intelligent system that uses advanced artificial intelligence technology, combined with historical patient data, to accurately assess the risk of ovarian cancer. The system constructs an artificial intelligence model to comprehensively analyze various data of patients, thereby predicting the occurrence risk of ovarian cancer, providing decision-making support for clinicians, helping to detect and intervene in ovarian cancer at an early stage, and improving the survival rate and quality of life of patients.

[0003] Traditional ovarian feature extraction methods have problems such as detail loss, inaccurate contour extraction, incomplete feature extraction, and insufficient data utilization; traditional ovarian cancer risk assessment model construction methods have problems such as a single kernel function, unreasonable penalty terms, incomplete objective functions, and unclear optimal conditions; traditional model performance improvement methods have problems such as unclear parameter performance evaluation, insufficient diversity maintenance, imperfect competition mechanisms, unreasonable compensation step sizes, and insufficient crossover mutations. Summary of the Invention

[0004] In view of the above situation, to overcome the defects of the prior art, the present invention provides an artificial intelligence-based ovarian cancer risk assessment assistance system. Aiming at the problems of detail loss, inaccurate contour extraction, incomplete feature extraction, and insufficient data utilization existing in traditional ovarian feature extraction methods, this solution improves the accuracy and comprehensiveness of feature extraction through detail enhancement, accurate contour extraction, comprehensive morphological feature calculation, texture feature calculation, feature integration, and data integration, providing richer and more effective data support for model training, thereby improving the accuracy and reliability of ovarian cancer risk assessment; aiming at the problems of a single kernel function, unreasonable penalty terms, incomplete objective functions, and unclear optimal conditions existing in traditional ovarian cancer risk assessment model construction methods, this solution improves the construction quality, classification performance, and stability of the model, and improves the accuracy and reliability of ovarian cancer risk assessment by constructing a double-layer fusion kernel function, designing innovative penalty terms, generating an objective function, and determining the optimal conditions; aiming at the problems of unclear parameter performance evaluation, insufficient diversity maintenance, imperfect competition mechanisms, unreasonable compensation step sizes, and insufficient crossover mutations existing in traditional model performance improvement methods, this solution improves the efficiency and stability of model performance improvement by constructing a diversity maintenance factor, designing a parameter competition mechanism, designing a compensation step size, and performing crossover mutations, thereby improving the accuracy and reliability of the ovarian cancer risk assessment model.

[0005] The technical solution adopted by the present invention is as follows: An ovarian cancer risk assessment assistance system based on artificial intelligence provided by the present invention includes a data acquisition module, an ovarian feature extraction module, a module for constructing an ovarian cancer risk assessment model, a module for improving model performance, and an ovarian cancer risk assessment assistance module;

[0006] The data acquisition module collects the clinical data, imaging data, laboratory test data, and final disease data of historical patients;

[0007] The ovarian feature extraction module performs ovarian feature extraction by detail enhancement, extracting the ovarian contour, morphological feature extraction, texture feature extraction, feature vector generation, and constructing a model training dataset;

[0008] The module for constructing an ovarian cancer risk assessment model constructs an ovarian cancer risk assessment model by setting labels, constructing a double-layer fusion kernel function, designing a penalty term, generating an objective function, and determining optimal conditions;

[0009] The module for improving model performance improves the model performance through initialization, parameter performance evaluation, constructing a diversity maintenance factor, designing a parameter competition mechanism, designing a compensation step size, crossover and mutation, and determining the process;

[0010] The ovarian cancer risk assessment assistance module uses the ovarian cancer risk assessment model to assist in assessing the ovarian cancer risk of patients.

[0011] Furthermore, the data acquisition module collects the clinical data, imaging data, laboratory test data, and final disease data of historical patients; the clinical data includes the age, family medical history, number of pregnancies, whether the abdomen aches, and whether there is abdominal distension of the patient; the imaging data refers to the ultrasonic examination images of the patient; the laboratory test data are the test results of tumor markers, blood routine, liver function, kidney function, and hormone levels; the final disease data refers to whether the patient has ovarian cancer.

[0012] Furthermore, the ovarian feature extraction module specifically includes the following contents:

[0013] Detail enhancement, performing weighted superposition and Laplacian operator enhancement on the original image, and combining linear gray-scale transformation and normalized detail highlighting through adaptive adjustment of parameters;

[0014] Extracting the ovarian contour, calculating the gradient amplitude and direction in the horizontal and vertical directions on the enhanced image, and selectively retaining the edge pixels consistent with the ovarian contour direction by using the edge intensity and direction information, thereby extracting the ovarian contour;

[0015] Morphological feature extraction, calculating the area, perimeter, major axis, and minor axis of the contour region, and further calculating the compactness and aspect ratio to quantitatively describe the regularity and stretching degree of the ovarian shape;

[0016] Texture feature extraction, calculating four texture metrics of contrast, correlation, energy, and homogeneity of the image based on the gray-level co-occurrence matrix to characterize the gray-level distribution and spatial relationship of the ovarian region;

[0017] Feature vector generation, forming an ovarian feature vector by combining the contrast, correlation, energy, homogeneity, compactness, and aspect ratio of the ovarian contour image;

[0018] Constructing a model training dataset, adding the patient's clinical data, laboratory test data, and final disease data to the ovarian feature vector to form a final data vector, creating a model training dataset, and adding the formed final data vector to the model training dataset.

[0019] Furthermore, the module for constructing the ovarian cancer risk assessment model specifically includes the following contents:

[0020] Setting labels, setting the final disease data as the label data of the ovarian cancer risk assessment model;

[0021] Constructing a double-layer fusion kernel function by fusing the Gaussian kernel function and the standardized part based on the inverse matrix of the covariance matrix;

[0022] Designing a penalty term, considering the spatial smoothness of the kernel function, introducing the weighted sum of the squares of the gradient norms to constrain the complexity of the model;

[0023] Generating an objective function, constructing the objective function by adding a quadratic penalty term and a kernel smoothing penalty while maximizing the Lagrangian dual objective;

[0024] Determining the optimal conditions, through the Lagrangian dual conditions, setting the gradients of the Lagrangian function with respect to the weights and biases to zero, and simultaneously satisfying the non-negativity of the Lagrange multipliers, the classification margin constraint, and the complementary slackness.

[0025] Furthermore, the module for improving model performance specifically includes the following contents:

[0026] Initialization, determining search parameters, including the bandwidth parameter and the scaling coefficient, and randomly generating initial parameter search points;

[0027] Parameter performance evaluation, setting the accuracy of the model as the parameter performance value and setting the parameter performance standard value;

[0028] Constructing a diversity maintenance factor, quantifying the breadth of the parameter distribution by the average value of the ratio of the range of the search point sets in each dimension to the corresponding search width to obtain the diversity maintenance factor;

[0029] Design a parameter competition mechanism to generate a competition probability by scaling the mean of the optimal and worst performance values with a diversity factor, and use this probability to weight the current, optimal, worst, and average positions to update the parameter point position;

[0030] Design a compensation step size. Calculate the compensation step size according to the ratio of the standard deviation of the search point performance to the average performance, and the difference between the average position of the initial point and the position of the point with the worst performance;

[0031] Cross - mutation. Adopt Gaussian perturbation and differential mutation strategies for parameter points, and use the compensation step size and randomly selected other search points to generate new candidate solutions;

[0032] Determine the process. Set the process of updating the parameter point position once using the parameter competition mechanism and then performing cross - mutation once as one search process; For each search, calculate the parameter performance value of the parameter search point after the search. If the parameter performance value is greater than the parameter performance standard value, the search stops and the parameter is qualified; Otherwise, the parameter is unqualified and the next search is carried out.

[0033] Furthermore, the ovarian cancer risk assessment auxiliary module first extracts ovarian characteristics by collecting the user's clinical data, imaging data, and laboratory test data, and then inputs the data after feature extraction into the ovarian cancer risk assessment auxiliary model. The model predicts the user's final disease status. If the prediction result is diseased, the ovarian cancer risk is high - risk; if the prediction result is not diseased, the ovarian cancer risk is low - risk.

[0034] The beneficial effects of the present invention using the above - mentioned scheme are as follows:

[0035] (1) Aiming at the problems of detail loss, inaccurate contour extraction, incomplete feature extraction, and insufficient data utilization existing in traditional ovarian feature extraction methods, this scheme improves the accuracy and comprehensiveness of feature extraction through detail enhancement, accurate contour extraction, comprehensive calculation of morphological features, texture feature calculation, feature integration, and data integration, providing richer and more effective data support for model training, thereby enhancing the accuracy and reliability of ovarian cancer risk assessment.

[0036] (2) Aiming at the problems of single kernel function, unreasonable penalty term, incomplete objective function, and unclear optimal conditions existing in traditional ovarian cancer risk assessment model construction methods, this scheme improves the construction quality, classification performance, and stability of the model, and enhances the accuracy and reliability of ovarian cancer risk assessment by constructing a double - layer fusion kernel function, designing an innovative penalty term, generating an objective function, and determining the optimal conditions.

[0037] (3) Aiming at the problems existing in the traditional model performance improvement methods, such as unclear parameter performance evaluation, insufficient diversity maintenance, imperfect competition mechanism, unreasonable compensation step, and insufficient crossover and mutation, this solution improves the efficiency and stability of model performance improvement by constructing a diversity maintenance factor, designing a parameter competition mechanism, designing a compensation step, and performing crossover and mutation, thereby improving the accuracy and reliability of the ovarian cancer risk assessment model. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 FIG. is a schematic diagram of an artificial intelligence-based ovarian cancer risk assessment assistance system provided by the present invention;

[0039] Figure 2 FIG. is a schematic diagram of an ovarian feature extraction module;

[0040] Figure 3 FIG. is a schematic diagram of a module for constructing an ovarian cancer risk assessment model;

[0041] Figure 4 FIG. is a schematic diagram of a model performance improvement module;

[0042] Figure 5 FIG. is a schematic diagram of the determination process in the model performance improvement module.

[0043] The drawings are used to provide a further understanding of the present invention and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the present invention, but do not constitute a limitation to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0044] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0045] In the description of the present invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc. indicating the orientation or positional relationship are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as a limitation to the present invention.

[0046] Example 1, refer to Figure 1, an ovarian cancer risk assessment assistance system based on artificial intelligence provided by the present invention includes a data acquisition module, an ovarian feature extraction module, an ovarian cancer risk assessment model construction module, a model performance improvement module, and an ovarian cancer risk assessment assistance module;

[0047] The data acquisition module collects the clinical data, imaging data, laboratory test data, and final disease data of historical patients, and sends the data to the ovarian feature extraction module;

[0048] The ovarian feature extraction module receives the data sent by the data acquisition module, and performs ovarian feature extraction by detail enhancement, ovarian contour extraction, morphological feature extraction, texture feature extraction, feature vector generation, and construction of a model training dataset, and sends the data to the ovarian cancer risk assessment model construction module;

[0049] The ovarian cancer risk assessment model construction module receives the data sent by the ovarian feature extraction module, and constructs an ovarian cancer risk assessment model by setting labels, constructing a double-layer fusion kernel function, designing a penalty term, generating an objective function, and determining optimal conditions, and sends the data to the model performance improvement module;

[0050] The model performance improvement module receives the data sent by the ovarian cancer risk assessment model construction module, and improves the model performance by initialization, parameter performance evaluation, construction of a diversity maintenance factor, design of a parameter competition mechanism, design of a compensation step size, crossover mutation, and determination of a process, and sends the data to the ovarian cancer risk assessment assistance module;

[0051] The ovarian cancer risk assessment assistance module receives the data sent by the model performance improvement module, and uses the ovarian cancer risk assessment model to assist in assessing the ovarian cancer risk of patients.

[0052] Example 2, refer to Figure 1 and Figure 2 , based on the above example, the ovarian feature extraction module specifically includes the following:

[0053] Detail enhancement, performing weighted superposition and Laplacian operator enhancement on the original image, and combining linear gray-scale transformation and normalized detail highlighting by adaptively adjusting parameters, as shown below:

[0054] ;

[0055] Among them, x and y respectively represent the abscissa index and ordinate index of the pixel point, represents the gray value of the enhanced image at the pixel point , , and are detail adaptation parameters, denotes the gray value of the original image at pixel ; and denote the minimum and maximum gray values in the image respectively, denotes the Laplacian operator of the image;

[0056] Extract the ovarian contour. Calculate the gradient magnitude and direction in the horizontal and vertical directions on the enhanced image. Use the edge intensity and direction information to selectively retain the edge pixels that are consistent with the ovarian contour direction, so as to extract the ovarian contour, which is expressed as follows:

[0057] ;

[0058] where denotes the edge intensity of the enhanced image at pixel ; denotes the partial derivative symbol, denotes taking the absolute value, denotes the edge direction, denotes the arctangent function, denotes the gray value of the ovarian contour image at pixel ; denotes the cosine function;

[0059] Morphological feature extraction. Calculate the area, perimeter, major axis and minor axis of the contour region, and then calculate the compactness and aspect ratio to quantitatively describe the regularity and stretching degree of the ovarian shape, which is expressed as follows:

[0060] ;

[0061] where , , , , and denote the compactness, area, perimeter, aspect ratio, major axis length and minor axis length of the ovarian contour image respectively, denotes pi;

[0062] Texture feature extraction. Calculate four texture metrics of contrast, correlation, energy and homogeneity of the image based on the gray-level co-occurrence matrix to characterize the gray distribution and spatial relationship of the ovarian region, which is expressed as follows:

[0063] ;

[0064] where , , and respectively represent the contrast, correlation, energy, and homogeneity of the image, where \(i\) and \(j\) represent the abscissa index and ordinate index of the gray-level co-occurrence matrix elements respectively, represents the value of the gray-level co-occurrence matrix at the position . represents the number of gray levels, and respectively represent the average gray levels of the image in the abscissa and ordinate directions;

[0065] Feature vector generation: The contrast, correlation, energy, homogeneity, compactness, and aspect ratio of the ovarian contour image are combined to form an ovarian feature vector, which is expressed as follows:

[0066] ;

[0067] where, represents the ovarian feature vector;

[0068] Construct a model training dataset: Add the patient's clinical data, laboratory test data, and final disease data to the ovarian feature vector to form a final data vector, create a model training dataset, and add the formed final data vector to the model training dataset.

[0069] By performing the above operations, aiming at the problems of detail loss, inaccurate contour extraction, incomplete feature extraction, and insufficient data utilization existing in the traditional ovarian feature extraction method, this solution improves the accuracy and comprehensiveness of feature extraction through detail enhancement, accurate contour extraction, comprehensive calculation of morphological features, texture feature calculation, feature integration, and data integration, provides richer and more effective data support for model training, and thus improves the accuracy and reliability of ovarian cancer risk assessment.

[0070] Example 3: Refer to Figure 1 and Figure 3 . Based on the above example, the module for constructing an ovarian cancer risk assessment model specifically includes the following content:

[0071] Set labels: Set the final disease data as the label data of the ovarian cancer risk assessment model;

[0072] Construct a double-layer fusion kernel function: Construct a double-layer fusion kernel function by fusing the Gaussian kernel function and the standardized part based on the inverse matrix of the covariance matrix, which is expressed as follows:

[0073] ;

[0074] where, and represent two data samples in the model training dataset, represents the double-layer fusion kernel function, represents the exponential function with the natural parameter as the base, represents the bandwidth parameter, represents taking the modulus length, represents the matrix transpose symbol, represents the inverse matrix of the covariance matrix of the model training dataset, represents the division by zero factor;

[0075] Design the penalty term. Considering the spatial smoothness of the kernel function, introduce the weighted sum of the squares of the gradient norms to constrain the complexity of the model, which is expressed as follows:

[0076] ;

[0077] where u represents the data sample variable, represents the weight of the model, represents the penalty term, q and v represent the indices of the data samples, M represents the total number of data samples, and respectively represent the data samples and 's Lagrange multipliers, and respectively represent the data samples and 's labels, represents the kernel function 's gradient with respect to u;

[0078] Generate the objective function. On the basis of maximizing the Lagrangian dual objective, add the quadratic penalty term and the kernel smoothing penalty at the same time to construct the objective function, which is expressed as follows:

[0079] ;

[0080] where a represents the Lagrange multiplier, represents taking the value of a when the whole expression reaches the maximum, and represent the scaling coefficients;

[0081] Determine the optimal conditions. Through the Lagrangian dual conditions, make the gradients of the Lagrangian function with respect to the weight and the bias equal to zero, and at the same time satisfy the non-negativity of the Lagrange multiplier, the classification margin constraint, and the complementary slackness, which is expressed as follows:

[0082] ;

[0083] where, represents the bias of the model, represents the Lagrangian function with respect to the model weight and the model bias The gradient is zero at the optimal solution, making the solution a stationary point in the parameter space; denotes for all, denotes that all Lagrange multipliers are non - negative; denotes the data sample mapping in the high - dimensional feature space, denotes that all samples are correctly classified and the margin is maximized; denotes that for each sample, either or holds, and only the support vectors contribute to the objective function.

[0084] By performing the above operations, aiming at the problems of single kernel function, unreasonable penalty term, incomplete objective function and unclear optimal conditions in the traditional ovarian cancer risk assessment model construction method, this solution improves the construction quality, classification performance and stability of the model by constructing a double - layer fusion kernel function, designing an innovative penalty term, generating an objective function and determining the optimal conditions, and enhances the accuracy and reliability of ovarian cancer risk assessment.

[0085] Example 4, refer to Figure 1 and Figure 4 and Figure 5 Based on the above - mentioned example, the model performance improvement module specifically includes the following:

[0086] Initialization, determine the search parameters, including the bandwidth parameter and the scaling coefficient, and randomly generate the initial parameter search points;

[0087] Parameter performance evaluation, set the accuracy of the model as the parameter performance value, and set the parameter performance standard value;

[0088] Construct the diversity maintenance factor, quantify the breadth of the parameter distribution through the average value of the ratio of the range of each - dimension search - point set to the corresponding search width, and obtain the diversity maintenance factor, which is expressed as follows:

[0089] ;

[0090] where, denotes the diversity maintenance factor, D denotes the total number of dimensions of the parameter search points, d denotes the dimension index of the parameter search points, denotes the search width of the parameter search points in the d - th dimension, denotes the maximum value of the searched parameter points in the d - th dimension, denotes the minimum value of the searched parameter points in the d - th dimension;

[0091] Design a parameter competition mechanism. Generate the competition probability by scaling the mean of the optimal and worst performance values with the diversity factor, and use this probability to weight the current, optimal, worst, and average positions to update the parameter point position, as shown below:

[0092] ;

[0093] where, represents the competition probability, represents the optimal parameter performance among all parameter search points, represents the worst parameter performance among all parameter search points, represents the parameter position after update, represents the parameter position before update, represents the optimal parameter search point among all parameter search points, represents the worst parameter search point among all parameter search points, represents the competition adjustment weight, represents the average position of all searched parameter points;

[0094] Design a compensation step size. Calculate the compensation step size according to the ratio of the standard deviation of the search point performance to the average performance, and the difference between the average position of the initial point and the position of the point with the worst performance, as shown below:

[0095] ;

[0096] where, represents the compensation step size, represents the average parameter performance of all searched parameter points, represents the standard deviation of the parameter performance of all searched parameter points, represents the average position of the initial parameter search points, represents the position of the parameter with the minimum parameter performance among the initial parameter search points;

[0097] Crossover and mutation. Adopt Gaussian perturbation and differential mutation strategies for parameter points, and use the compensation step size and other randomly selected search points to generate new candidate solutions, as shown below:

[0098] ;

[0099] where, represents the position of the parameter search point after crossover and mutation, represents the position of the parameter search point before crossover and mutation, 、 and represent the mutation weights, represents a random number obeying the standard normal distribution, r1 represents a random number with a value range between 0 and 1, , and represent the positions of the parameter search points randomly selected from all parameter search points;

[0100] Determination process: Set the process of updating the position of the parameter point once using the parameter competition mechanism and then performing one crossover and mutation as one search process; For each search, calculate the parameter performance value of the parameter search point after the search. If the parameter performance value is greater than the parameter compliance value, the search stops and the parameter is qualified; Otherwise, the parameter is unqualified and the next search is performed.

[0101] By performing the above operations, aiming at the problems of unclear parameter performance evaluation, insufficient diversity maintenance, imperfect competition mechanism, unreasonable compensation step length, and insufficient crossover and mutation existing in the traditional model performance improvement method, this solution improves the efficiency and stability of model performance improvement by constructing a diversity maintenance factor, designing a parameter competition mechanism, designing a compensation step length, and performing crossover and mutation, thereby improving the accuracy and reliability of the ovarian cancer risk assessment model.

[0102] Example Five. Refer to Figure 1 , this example is based on the above example. The ovarian cancer risk assessment auxiliary module collects the user's clinical data, imaging data, and laboratory test data, first performs ovarian feature extraction, and then inputs the data after feature extraction into the ovarian cancer risk assessment auxiliary model. The model predicts the user's final disease status. If the prediction result is diseased, the ovarian cancer risk is high risk; if the prediction result is not diseased, the ovarian cancer risk is low risk.

[0103] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0104] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention.

[0105] The above description of the present invention and its implementation manners is not restrictive. What is shown in the drawings is only one of the implementation manners of the present invention, and the actual structure is not limited thereto. In general, if those of ordinary skill in the art are inspired by it and, without departing from the gist of the present invention, design similar structural manners and embodiments to this technical solution without creative efforts, they shall fall within the protection scope of the present invention.

Claims

1. An artificial intelligence-based ovarian cancer risk assessment assistance system, characterized in that: It includes a data acquisition module, an ovarian feature extraction module, a module for constructing an ovarian cancer risk assessment model, a model performance improvement module, and an ovarian cancer risk assessment assistance module; The data acquisition module acquires the clinical data, imaging data, laboratory test data, and final disease data of historical patients; The ovarian feature extraction module performs ovarian feature extraction by detail enhancement, extracting the ovarian contour, morphological feature extraction, texture feature extraction, feature vector generation, and constructing a model training dataset; The module for constructing an ovarian cancer risk assessment model constructs an ovarian cancer risk assessment model by setting labels, constructing a double-layer fusion kernel function, designing a penalty term, generating an objective function, and determining the optimal conditions; The model performance improvement module improves the model performance through initialization, parameter performance evaluation, constructing a diversity maintenance factor, designing a parameter competition mechanism, designing a compensation step size, crossover mutation, and determining the process; The ovarian cancer risk assessment assistance module uses the ovarian cancer risk assessment model to assist in assessing the ovarian cancer risk of patients.

2. The ovarian cancer risk assessment assistance system based on artificial intelligence according to claim 1, wherein: The ovarian feature extraction module specifically includes the following: Detail enhancement, performing weighted superposition and Laplacian operator enhancement on the original image, and combining linear gray-level transformation with normalized detail highlighting by adaptively adjusting parameters; Extracting the ovarian contour, calculating the gradient magnitude and direction in the horizontal and vertical directions on the enhanced image, and selectively retaining the edge pixels consistent with the ovarian contour direction using the edge intensity and direction information to extract the ovarian contour; Morphological feature extraction, calculating the area, perimeter, major axis, and minor axis of the contour region, and further calculating the compactness and aspect ratio to quantitatively describe the regularity and stretching degree of the ovarian shape; Texture feature extraction, calculating four texture metrics of contrast, correlation, energy, and homogeneity of the image based on the gray-level co-occurrence matrix to characterize the gray-level distribution and spatial relationship of the ovarian region; Feature vector generation, forming an ovarian feature vector by combining the contrast, correlation, energy, homogeneity, compactness, and aspect ratio of the ovarian contour image; Constructing a model training dataset, adding the clinical data, laboratory test data, and final disease data of the patient to the ovarian feature vector to form a final data vector, creating a model training dataset, and adding the formed final data vector to the model training dataset.

3. The ovarian cancer risk assessment assistance system based on artificial intelligence according to claim 1, wherein: The module for constructing an ovarian cancer risk assessment model specifically includes the following: Setting labels, setting the final disease data as the label data of the ovarian cancer risk assessment model; Constructing a double-layer fusion kernel function, constructing a double-layer fusion kernel function by fusing the Gaussian kernel function and the standardized part based on the inverse matrix of the covariance matrix; Designing a penalty term, considering the spatial smoothness of the kernel function, introducing the weighted sum of the squares of the gradient norms to constrain the complexity of the model; Generating an objective function, constructing an objective function by adding a quadratic penalty term and a kernel smoothing penalty while maximizing the Lagrangian dual objective; Determining the optimal conditions, through the Lagrangian dual conditions, making the gradients of the Lagrangian function with respect to the weights and biases zero, and simultaneously satisfying the non-negativity of the Lagrange multipliers, the classification margin constraint, and the complementary slackness.

4. An ovarian cancer risk assessment assistance system based on artificial intelligence according to claim 1, characterized in that: The model performance improvement module specifically includes the following: Initialization, determine search parameters, including bandwidth parameter and scaling factor, and randomly generate an initial parameter search point; Parameter performance evaluation, set the accuracy of the model as the parameter performance value, and set the parameter performance compliance value; Construct a diversity maintenance factor, quantify the breadth of parameter distribution by the average value of the ratio of the range of the search point set in each dimension to the corresponding search width, and obtain the diversity maintenance factor; Design a parameter competition mechanism, generate a competition probability by scaling the mean of the optimal and worst performance values by the diversity factor, and update the parameter point position by weighting the current, optimal, worst, and average positions with this probability; Design a compensation step size, calculate the compensation step size according to the ratio of the standard deviation of the search point performance to the average performance, and the difference between the average position of the initial point and the position of the point with the worst performance; Crossover and mutation, adopt Gaussian perturbation and differential mutation strategies for parameter points, and generate new candidate solutions using the compensation step size and other randomly selected search points; Determine the process, set the process of updating the position of the parameter point once using the parameter competition mechanism and then performing one crossover and mutation as one search process; for each search, calculate the parameter performance value of the parameter search point after the search. If the parameter performance value is greater than the parameter performance compliance value, the search stops and the parameter is qualified; otherwise, the parameter is unqualified and the next search is performed.

5. An ovarian cancer risk assessment assistance system based on artificial intelligence according to claim 1, characterized in that: The data acquisition module acquires the clinical data, imaging data, laboratory test data, and final disease data of historical patients; the clinical data includes the age, family history, number of pregnancies, whether there is abdominal pain and whether there is abdominal distension of the patient; the imaging data refers to the ultrasound examination images of the patient; the laboratory test data are the test results of tumor markers, blood routine, liver function, kidney function, and hormone levels; the final disease data refers to whether the patient has ovarian cancer; The ovarian cancer risk assessment assistance module acquires the clinical data, imaging data, and laboratory test data of the user, first performs ovarian feature extraction, and then inputs the data after feature extraction into the ovarian cancer risk assessment assistance model. The model predicts the final disease condition of the user. If the prediction result is diseased, the ovarian cancer risk is high risk; if the prediction result is not diseased, the ovarian cancer risk is low risk.

Images