Model for predicting risk of ovarian function decline

Abstract

The invention is applied to the technical field of risk prediction models, and particularly discloses a model for predicting ovarian function decline risk, the prediction model comprises risk factors, and the risk factors calculate risk probability according to a logistic regression equation. According to the model for predicting the ovarian function decline risk, a plurality of risk factors influencing the ovarian function are introduced on the basis of a logistic regression equation, the ovarian function prediction model is constructed according to the linear regression relation between the risk factors and the ovarian function and assisted by a large number of clinical data samples, and the ovarian function decline risk is predicted through calculation of the prediction model. According to the method, the risk score and the risk probability of group ovarian function decline under different conditions can be obtained, a data sample is substituted into the prediction model, the obtained result is compared with the data sample, various parameters in the prediction model are continuously corrected and iterated according to the comparison result, and the accuracy and the stability of the prediction model are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of risk prediction models, and in particular to a model for predicting the risk of ovarian function decline. Background Art

[0002] The ovaries are located in the female pelvic cavity and are paired solid organs. The ovaries are female gonads and are flat and oval in shape. The main function of the ovaries is to produce and discharge eggs and secrete sex hormones to promote the development and maintenance of female sexual characteristics. Generally speaking, the left and right ovaries discharge a mature egg alternately every month. Ovarian function reflects a woman's reproductive endocrine function. Ovarian dysfunction is common in the reproductive period due to diseases such as decreased ovarian reserve function and premature ovarian failure. These diseases seriously affect women's pregnancy rate. There are many factors that affect ovarian function, including age, whether there is a history of genetic diseases, lifestyle habits, etc. Various factors act together on the ovaries and affect ovarian function.

[0003] At present, there are usually several types of clinical evaluation methods for ovarian function. Ovarian function is evaluated by measuring the levels of follicle-stimulating hormone (FSH), luteinizing hormone (LH), estradiol (E2), anti-Mullerian hormone (AMH) and other hormones in serum; ovarian size and antral follicle count are measured by transvaginal ultrasound. However, clinical detection methods have certain fluctuations in the determination of ovarian function. The results of a single test sometimes cannot fully and accurately reflect the long-term functional status of the ovaries. Multiple tests may be required for comprehensive judgment. In addition, the detection process is relatively cumbersome, which is not conducive to quickly judging the ovarian function according to the patient's condition. Summary of the invention

[0004] The purpose of the present invention is to provide a model for predicting the risk of decreased ovarian function, so as to solve the problem in the above-mentioned background technology that the process of clinical testing and evaluating ovarian function is relatively cumbersome, thus making it inconvenient to quickly judge ovarian function according to one's own situation.

[0005] To achieve the above object, the present invention provides the following technical solutions: a model for predicting the risk of ovarian function decline, the prediction model includes risk factors, and the risk probability of the risk factors is calculated according to a logistic regression equation; Preferably, the risk factors include age, body mass index, genetic disease history, iatrogenic injury, endocrine hormones, medication history, autoimmune factors and alloimmune factors, and the above risk factors are marked as X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X12, X13, X14, X15, X16, X17, X18, X19, X20, X21, X30, X41, X50, X61, X71, X80, X9, X10 6， X7,X 8。 The corresponding relationship between the risk factors and markers is as follows: age X1, body mass index X2, genetic disease history X3, iatrogenic injury X4, endocrine hormone X5 5、 Medication history X6, autoimmune factors X7 and alloimmune factors X8.

[0006] The above technical solution is used to perform multi-parameter prediction of ovarian function by setting multiple risk factors, ensuring that the prediction model is specific to different individuals.

[0007] Preferably, the coefficients of each risk factor and the constant of the logistic regression of the prediction model are estimated using SPSS, and the calculated coefficients of the risk factors are marked as age β1, body mass index β2, genetic disease history β3, iatrogenic injury β4, endocrine hormone β5, medication history β6, autoimmune factors β7 and alloimmune factors β8, and the constant of the logistic regression is marked as K.

[0008] By adopting the above technical solution, a calculation formula of the prediction model is constructed based on the logistic regression equation, and the prediction model is used to predict the ovarian function under the influence of multiple parameters.

[0009] Preferably, a benchmark risk value D is set for each risk factor, starting from the benchmark risk value and ending at D i The risk factors are incremented to obtain the risk score of each risk factor, which is recorded as C. The risk score of the baseline risk value D is recorded as 0. i The incremental gradient risk score is increased by 1.

[0010] By adopting the above technical solution, the risk score of each risk factor can be determined using the benchmark risk value D, and the degree of ovarian function decline can be calculated based on the risk score.

[0011] Preferably, the risk score of each risk factor in the prediction model is calculated by the following formula: The risk factor X i The risk score C is calculated by the following formula: The risk probability P of ovarian function decline in the prediction model is calculated by the following formula: .

[0012] By adopting the above technical solution and using various calculation formulas in the prediction model, the risk of ovarian function decline under different individual conditions and the probability of ovarian function decline under various conditions can be calculated.

[0013] Preferably, the steps of constructing and maintaining the prediction model are as follows: Data collection and processing steps: Based on clinical case data, collect enough ovarian data samples, clean the collected sample data, deal with missing values, outliers and erroneous data, and use 75% of the processed sample data to build the model and 25% to evaluate the training model; Parameter determination step: Use 75% of the sample data to calculate the coefficient β of each risk factor i , and the logistic regression constant K, are estimated based on clinical data to determine the baseline risk value D and incremental gradient D of each risk factor i ; Model substitution step: Substitute the above parameters into the calculation formula of the prediction model; Model evaluation and training steps: Calculate the probability of ovarian function decline using the prediction model for each risk factor in 25% of the sample data, and compare and analyze the results with the sample data; Model optimization steps: Modify the parameters of the prediction model according to the comparison results, and then use the sample data again for repeated evaluation and training.

[0014] By using the above technical solution, the constructed prediction model can be evaluated and trained using sample data. By comparing the prediction model results with the sample data, various parameters in the prediction model can be corrected to ensure the accuracy of the prediction model.

[0015] Compared with the prior art, the beneficial effects of the present invention are: the model for predicting the risk of ovarian function decline: 1. In the present invention, based on the logistic regression equation, multiple risk factors affecting ovarian function are introduced, and according to the linear regression relationship between risk factors and ovarian function, a large number of clinical data samples are used to construct an ovarian function prediction model. Through the calculation of the prediction model, the risk score and risk probability of ovarian function decline in groups under different conditions can be obtained. The data samples are substituted into the prediction model, and the obtained results are compared with the data samples. Through the comparison results, various parameters in the prediction model are continuously revised and iterated to increase the accuracy and stability of the prediction model; 2. The prediction model in the present invention enables patients to quickly understand their own ovarian health risk status, enhance self-care awareness, actively adopt a healthy lifestyle and regular check-ups, assist in clinical decision-making, and help discover potential ovarian lesion risks in the early stages of the disease or even before symptoms appear, providing possibilities for early intervention and treatment. The prediction model is set up with multiple risk factor influences, so that the prediction model can provide personalized prevention and monitoring recommendations based on individual specific risk factors, such as genetic factors, lifestyle, hormone levels, etc., to improve the pertinence and effectiveness of medical treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic diagram of the maintenance process structure is constructed for the model of the present invention; DETAILED DESCRIPTION The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0017] See also Figure 1 The present invention provides a technical solution: a model for predicting the risk of decreased ovarian function.

[0018] The prediction model includes risk factors, which include age, body mass index, genetic disease history, iatrogenic injury, endocrine hormones, medication history, autoimmune factors and alloimmune factors. The above risk factors are marked as X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X12, X13, X14, X15, X16, X17, X18, X19, X20, X21, X30, X41, X50, X61, X71, X8 6， X7,X 8。 The corresponding relationship between the risk factors and markers is as follows: age X1, body mass index X2, genetic disease history X3, iatrogenic injury X4, endocrine hormone X5 5、 Medication history X6, autoimmune factors X7 and alloimmune factors X8.

[0019] Ovarian function is affected by many factors. The prediction model in the present invention selects 8 representative risk factors to analyze their effects on ovarian function. The effect of risk factor age X1 on ovarian function is manifested in that ovarian function changes significantly with age. In the age group of 20s to 30s, ovarian function is usually in a relatively vigorous period. From around 35 years old, ovarian function gradually declines. After 40 years old, the decline of ovarian function accelerates. The effect of body mass index X2 on ovarian function is manifested in that the body's fat content affects the synthesis of estrogen, thereby affecting ovarian function. Genetic history X3, iatrogenic injury X4, endocrine hormones X5, medication history X6, autoimmune factors X7 and alloimmune factors X8 will all affect hormones in the human body, thereby affecting ovarian function. The personalized prediction of human ovarian function is achieved through the influence of multiple factors on ovarian function.

[0020] The coefficients of each risk factor and the constant of the logistic regression of the prediction model were estimated using SPSS. The coefficients of the calculated risk factors were marked as age β1, body mass index β2, genetic disease history β3, iatrogenic injury β4, endocrine hormone β5, medication history β6, autoimmune factors β7, alloimmune factors β8, and the constant of logistic regression was marked as K; The prediction model in the present invention is based on the logistic regression equation, and the multi-factor risk factors are used to jointly construct the prediction model calculation formula. The coefficients of the risk factors are calculated using SPSS software. The coefficient of each risk factor is calculated using SPSS software and is recorded as β i A large amount of ovarian sample data was input into SPSS software, and the sample data was processed and analyzed according to SPSS software to obtain the constant of logistic regression, denoted as K, which was convenient for subsequent model construction. At the same time, the coefficient of each risk factor was calculated, and they were marked as age β1, body mass index β2, genetic disease history β3, iatrogenic injury β4, endocrine hormone β5 and medication history β6, autoimmune factors β7, and alloimmune factors β8.

[0021] For each risk factor, a benchmark risk value D is set. i The risk factors are incremented to calculate the risk score of each risk factor. The risk score is recorded as C, and the risk score of the baseline risk value D is recorded as 0. i The incremental gradient risk score is increased by 1; The risk score C is calculated for different risk factors. Taking age X1 as an example, 35 years old is used as the benchmark risk value D1. The risk score C for individuals under 35 years old is i 0, with 5 years as one D i Increasing gradient, the risk score is recorded as C, and the calculation result is rounded up. For example, the risk score C of a 41-year-old individual sample is 3. The risk scores C of different sample individuals are calculated by analogy. The risk scores C of the remaining binary risk factors are recorded as 0 or 1, such as whether there is a history of genetic disease X3. If there is a history of genetic disease X3, the risk score C is recorded as 1, and if there is no history of genetic disease X3, the risk score C is recorded as 0. And so on, the risk scores C of all risk factors are obtained by analogy.

[0022] The risk score of each risk factor in the prediction model is calculated by the following formula: Among them, risk factor X i The risk score C is calculated by the following formula: The risk probability P of ovarian function decline in the prediction model is calculated by the following formula: ; The risk score of each risk factor is calculated according to the above formula, and the sum of the risk scores of all risk factors is calculated using the sum formula. The larger the data is, the greater the risk of ovarian function decline of the individual. The probability of ovarian function decline of the individual can be calculated using the risk probability P calculation formula, which is convenient for predicting ovarian function of different individuals.

[0023] The steps for building and maintaining the prediction model are as follows: Data collection and processing step: Collect enough ovarian data samples based on clinical case data, clean the collected sample data, handle missing values, outliers and erroneous data, and use 75% of the processed sample data to build the model and 25% to evaluate the training model; Parameter determination step: Use 75% of the sample data to calculate the coefficient β of each risk factor i , and the logistic regression constant K, are estimated based on clinical data to determine the baseline risk value D and incremental gradient D of each risk factor i ; Model substitution step: Substitute the above parameters into the calculation formula of the prediction model; Model evaluation and training step: Use the prediction model to calculate the probability of ovarian function decline for each risk factor in 25% of the sample data, and compare and analyze the results with the sample data; Model optimization step: Modify the various parameters of the prediction model according to the comparison results, and use the sample data again for repeated evaluation and training after modification; like Figure 1 As shown, clinical case data were collected, a large number of ovarian data samples were processed, the sample data were cleaned, missing values, outliers and erroneous data were processed, and the processed data samples were used to construct the prediction model. 75% of them were used to build the model, and the coefficient β of each risk factor was calculated using SPSS. i , and the logistic regression constant K, coefficient β i After the constant K is calculated, the baseline risk value D and the incremental gradient D of each risk factor are estimated based on clinical data. i ; Substitute each parameter into the calculation formula of the prediction model. At this time, the data of different individuals can be substituted into the calculation formula to calculate the risk value of ovarian function decline and the risk probability of ovarian function decline.

[0024] The remaining 25% of the sample data were substituted into the prediction model to calculate the probability of ovarian function decline, and the result was compared and analyzed with the sample data. The comparison results were used to correct the various parameters of the prediction model. After correction, the sample data were used again for repeated evaluation and training, thereby improving the prediction accuracy of the prediction model.

[0025] Working principle: In the present invention, the prediction model is based on the logistic regression equation, which introduces multiple parameters that affect ovarian function, including age X1, body mass index X2, genetic disease history X3, iatrogenic injury X4, endocrine hormones X5 and medication history X6, autoimmune factors X7, and 7， Alloimmune Factor X 8。According to the linear regression relationship between risk factors and ovarian function, SPSS and a large number of clinical data samples were used to calculate the coefficient β of each risk factor. i , and the logistic regression constant K, estimate the baseline risk value D and incremental gradient D of each risk factor i ; Calculate the risk score C, substitute the parameters into the calculation formula, and construct an ovarian function prediction model. Through the calculation of the prediction model, the risk score and risk probability of ovarian function decline in groups under different conditions can be obtained. At the same time, the data sample can be substituted into the prediction model, and the obtained result can be compared with the data sample. Through the comparison results, the various parameters in the prediction model are continuously revised and iterated to increase the accuracy and stability of the prediction model. The use of the prediction model is helpful to discover potential ovarian lesion risks in the early stages of the disease or even before symptoms appear, and the prediction model is set with multiple risk factors, so that the prediction model can provide personalized prevention and monitoring suggestions based on the individual's specific risk factors, thereby improving the pertinence and effectiveness of medical treatment.

[0026] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A model for predicting the risk of ovarian function decline, characterized in that: The prediction model includes risk factors, and the risk probability of the risk factors is calculated according to the logistic regression equation.

2. A model for predicting the risk of ovarian function decline according to claim 1, characterized in that: Risk factors include age, body mass index, genetic disease history, iatrogenic injury, endocrine hormones, medication history, autoimmune factors and alloimmune factors. The above risk factors are marked as X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X12, X13, X14, X15, X16, X17, X18, X19, X20, X21, X30, X41, X50, X61, X71, X8 6， X7,X 8。 The corresponding relationship between the risk factors and markers is as follows: age X1, body mass index X2, genetic disease history X3, iatrogenic injury X4, endocrine hormone X5 5、 Medication history X6, autoimmune factors X7 and alloimmune factors X8.

3. A model for predicting the risk of ovarian function decline according to claim 1, characterized in that: The coefficients of each risk factor and the constant of the logistic regression of the prediction model were estimated using SPSS. The calculated coefficients of the risk factors were marked as age β1, body mass index β2, genetic disease history β3, iatrogenic injury β4, endocrine hormone β5, medication history β6, autoimmune factors β7 and alloimmune factors β8, and the constant of the logistic regression was marked as K.

4. A model for predicting the risk of ovarian function decline according to claim 1, characterized in that: A benchmark risk value D is set for each risk factor, starting from the benchmark risk value and ending at D i The risk factors are incremented to obtain the risk score of each risk factor, which is recorded as C. The risk score of the baseline risk value D is recorded as 0. i The incremental gradient risk score is increased by 1.

5. A model for predicting the risk of ovarian function decline according to claim 1, characterized in that: The risk score of each risk factor in the prediction model is calculated by the following formula: The risk factor X i The risk score C is calculated by the following formula: The risk probability P of ovarian function decline in the prediction model is calculated by the following formula: 。 6. A model for predicting the risk of ovarian function decline according to claim 1, characterized in that: The steps for building and maintaining the prediction model are as follows: Data collection and processing steps: Based on clinical case data, collect enough ovarian data samples, clean the collected sample data, deal with missing values, outliers and erroneous data, and use 75% of the processed sample data to build the model and 25% to evaluate the training model; Parameter determination step: Use 75% of the sample data to calculate the coefficient β of each risk factor i , and the logistic regression constant K, are estimated based on clinical data to determine the baseline risk value D and incremental gradient D of each risk factor i ; Model substitution step: Substitute the above parameters into the calculation formula of the prediction model; Model evaluation and training steps: Calculate the probability of ovarian function decline using the prediction model for each risk factor in 25% of the sample data, and compare and analyze the results with the sample data; Model optimization steps: Modify the parameters of the prediction model according to the comparison results, and then use the sample data again for repeated evaluation and training.

Images