Ovarian response level and cycle cancellation risk prediction method, system, equipment, medium and product

Abstract

The invention discloses an ovarian response level and periodic cancel risk prediction method, system, device, medium and product, and relates to the field of medical models.The method comprises the steps that a prediction model including an egg obtaining number prediction model, a periodic cancel risk prediction model and an ovarian response level prediction model is constructed; acquiring clinical characteristic variables of the COS patient; constructing a feature data set based on the clinical feature variables; inputting the feature data set into a prediction model to obtain a prediction result; obtaining a prediction value based on the prediction result; the prediction value comprises an ovarian response level prediction value, a period cancel risk prediction value and an ovum obtaining number prediction value. According to the application, the ovarian response level under different exogenous Gn usage levels or the risk of canceling fresh transplantation due to the hyperstimulus syndrome can be evaluated according to the condition when the patient begins to accept the exogenous Gn.

Description

Technical Field

[0001] The present application relates to the field of medical models, and particularly to a method, system, device, medium and product for predicting ovarian response level and cycle cancellation risk. Background Art

[0002] Controlled ovarian stimulation (COS) is a key link in the treatment of assisted reproductive technologies (ART). The ovarian response level during the COS process is closely related to the quantity and quality of eggs. However, too high an ovarian response level may also cause iatrogenic diseases such as ovarian hyperstimulation syndrome and lead to the cancellation of fresh embryo transfer. Therefore, it is very important to maintain an appropriate ovarian response level during the COS process. Clinically, the ovarian response level is generally controlled by adjusting the dosage of exogenous gonadotropin (Gn).

[0003] The method of adjusting the dosage of exogenous Gn is usually based on clinical experience or a model based on a certain "target number of retrieved eggs". Clinical experience is usually subjectively evaluated by clinicians according to information such as the patient's age, physique, reproductive endocrinology, and follicular ultrasound, and the dosage of exogenous Gn is determined and adjusted in combination with the treatment situation, which has the disadvantages of poor repeatability and high uncertainty. The "target number of retrieved eggs" model is to reverse the dosage of exogenous Gn required to reach the target number of retrieved eggs according to the specified "target number of retrieved eggs" and the regression model of past data. This model cannot provide an estimate of the risks at different dosages of exogenous Gn. These methods cannot evaluate the clinical risks at different levels of exogenous Gn usage. These risks usually include: the risk of hyperstimulation syndrome or cancellation of fresh embryo transfer due to hyperstimulation syndrome and the risk of reduced ovarian response. Therefore, there is an urgent need for a method for predicting ovarian response level and cycle cancellation risk to evaluate the ovarian response level or the risk of cancellation of fresh embryo transfer due to hyperstimulation syndrome at different levels of exogenous Gn usage. Summary of the Invention

[0004] The purpose of the present application is to provide a method, system, device, medium and product for predicting ovarian response level and cycle cancellation risk. It can evaluate the ovarian response level or the risk of cancellation of fresh embryo transfer due to hyperstimulation syndrome at different levels of exogenous Gn usage according to the situation when the patient starts to receive exogenous Gn.

[0005] To achieve the above purpose, the present application provides the following solutions:

[0006] In the first aspect, the present application provides a method for predicting ovarian response level and cycle cancellation risk, including:

[0007] Construct a prediction model; the prediction model includes: an egg retrieval number prediction model, a cycle cancellation risk prediction model, and an ovarian response level prediction model;

[0008] Obtain the clinical characteristic variables of COS patients; the clinical characteristic variables include: female age, ovulation induction protocol, previous ovarian hyporesponse, previous ovarian hyperresponse, infertility duration, thyroid disorder diagnosis, polycystic ovary syndrome, endometriosis, age at menarche, menstrual period longer than A days, menstrual period shorter than B days, heavy menstrual flow, menstrual cycle shorter than C days, menstrual cycle longer than D days, amenorrhea, irregular menstrual cycle, height, weight, body mass index, basal follicle-stimulating hormone, basal luteinizing hormone, basal prolactin, basal estrogen, basal progesterone, basal testosterone, basal anti-Müllerian hormone, left basal antral follicles, right basal antral follicles, total antral follicle count, starting dose, starting timing, follicle-stimulating hormone on the starting day, luteinizing hormone on the starting day, estrogen on the starting day, AMH on the starting day, human menopausal gonadotropin HMG, recombinant FSH, urinary FSH, follow-up interval, ovarian cysts, mean follicle diameter on the starting day, number of follicles on the starting day, and standard deviation of follicle diameter on the starting day;

[0009] Construct a feature dataset based on the clinical characteristic variables;

[0010] Input the feature dataset into the prediction model to obtain a prediction result; the prediction result includes an ovarian response level prediction result, a cycle cancellation risk prediction result, and an egg retrieval number prediction result;

[0011] Obtain a predicted value based on the prediction result; the predicted value includes an ovarian response level predicted value, a cycle cancellation risk predicted value, and an egg retrieval number predicted value.

[0012] In a second aspect, the present application provides a system for predicting ovarian response level and cycle cancellation risk, including:

[0013] A data acquisition module for obtaining the clinical characteristic variables of COS patients;

[0014] A data processing module connected to the data acquisition module for obtaining a feature dataset based on the clinical characteristic variables;

[0015] A data prediction module connected to the data processing module for obtaining a prediction result using a prediction model based on the feature dataset and obtaining a predicted value based on the prediction result;

[0016] A display module connected to the data prediction module for displaying the predicted value.

[0017] In a third aspect, the present application provides a computer device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the ovarian response level and cycle cancellation risk prediction method described in any one of the above.

[0018] In a fourth aspect, the present application provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the ovarian response level and cycle cancellation risk prediction method described in any one of the above are implemented.

[0019] In a fifth aspect, the present application provides a computer program product, including a computer program, and when the computer program is executed by a processor, the steps of the ovarian response level and cycle cancellation risk prediction method described in any one of the above are implemented.

[0020] According to the specific embodiments provided by the present application, the present application has the following technical effects:

[0021] The present application provides a method, system, device, medium and product for predicting ovarian response level and cycle cancellation risk. By constructing a prediction model including an oocyte retrieval number prediction model, a cycle cancellation risk prediction model and an ovarian response level prediction model, and based on this prediction model and the clinical characteristic variables of COS patients, the oocyte retrieval number, cycle cancellation risk and ovarian response level are respectively predicted to obtain predicted values, and the ovarian response level under different exogenous Gn usage levels or the risk of canceling fresh transplantation due to ovarian hyperstimulation syndrome can be evaluated according to the situation when the patient starts to receive exogenous Gn. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application, and those of ordinary skill in the art can obtain other drawings without creative efforts based on these drawings.

[0023] Figure 1 It is a flowchart of a method for predicting ovarian response level and cycle cancellation risk in an embodiment of the present application;

[0024] Figure 2 It is a scatter plot of the predicted oocyte retrieval number and the observed value provided in an embodiment of the present application;

[0025] Figure 3 It is an ROC curve graph for predicting cycle cancellation risk provided in an embodiment of the present application;

[0026] Figure 4Calibration curve graph for predicting the risk of cycle cancellation provided by an embodiment of the present application;

[0027] Figure 5 Multi-class macro ROC curve graph for predicting ovarian response level provided by an embodiment of the present application;

[0028] Figure 6 Multi-class micro ROC curve graph for predicting ovarian response level provided by an embodiment of the present application;

[0029] Figure 7 ROC curve graph for predicting low ovarian response level provided by an embodiment of the present application;

[0030] Figure 8 ROC curve graph for predicting appropriate ovarian response level provided by an embodiment of the present application;

[0031] Figure 9 ROC curve graph for predicting high ovarian response level provided by an embodiment of the present application;

[0032] Figure 10 Calibration curve graph for predicting low ovarian response level provided by an embodiment of the present application;

[0033] Figure 11 Calibration curve graph for predicting appropriate ovarian response level provided by an embodiment of the present application;

[0034] Figure 12 Calibration curve graph for predicting high ovarian response level provided by an embodiment of the present application;

[0035] Figure 13 Comparison graph of the egg retrieval number prediction results of the prediction model of the present application and the existing prediction model for predicting the egg retrieval number of the same population provided by an embodiment of the present application;

[0036] Figure 14 Structural schematic diagram of a computer device provided by an embodiment of the present application. Detailed implementation manners

[0037] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0038] To make the above objects, features, and advantages of the present application more obvious and understandable, the present application will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.

[0039] In an exemplary embodiment, as Figure 1 shown, a method for predicting ovarian response level and cycle cancellation risk is provided. This method is executed by a computer device, and specifically, it can be executed alone by a computer device such as a terminal or a server, or jointly executed by a terminal and a server. In the embodiments of the present application, taking the prediction of ovarian response level and cycle cancellation risk of COS patients by this method as an example, it includes:

[0040] Step 1, construct a prediction model. The prediction model includes: an oocyte retrieval number prediction model, a cycle cancellation risk prediction model, and an ovarian response level prediction model.

[0041] Specifically, the process of constructing the prediction model includes:

[0042] Obtain historical COS patient data at different exogenous Gn usage levels; the historical COS patient data includes clinical characteristic variables of historical COS patients and corresponding observed values. Construct a training data set based on the clinical characteristic variables of historical COS patients. Construct a validation data set based on the historical COS patient data. Construct an initial prediction model based on the training data set. Calibrate the initial prediction model based on the validation data set to obtain the prediction model.

[0043] Step 2, obtain the clinical characteristic variables of COS patients.

[0044] Specifically, the clinical characteristic variables of COS patients include: female age, ovulation induction protocol, previous ovarian hyporesponse, previous ovarian hyperresponse, infertility duration, thyroid disorder diagnosis, polycystic ovary syndrome, endometriosis, age at menarche, menstrual period longer than A days, menstrual period shorter than B days, heavy menstrual flow, menstrual cycle shorter than C days, menstrual cycle longer than D days, amenorrhea, irregular menstrual cycle, height, weight, body mass index, basal follicle-stimulating hormone, basal luteinizing hormone, basal prolactin, basal estrogen, basal progesterone, basal testosterone, basal anti-Müllerian hormone, left basal antral follicles, right basal antral follicles, total antral follicle count, starting dose, starting timing, follicle-stimulating hormone on the starting day, luteinizing hormone on the starting day, estrogen on the starting day, AMH on the starting day, human menopausal gonadotropin HMG, recombinant FSH, urinary FSH, follow-up interval, ovarian cysts, mean follicle diameter on the starting day, number of follicles on the starting day, and standard deviation of follicle diameter on the starting day.

[0045] Step 3, construct a feature data set based on the clinical characteristic variables.

[0046] Step 4, input the feature data set into the prediction model to obtain a prediction result. The prediction result includes an ovarian response level prediction result, a cycle cancellation risk prediction result, and an oocyte retrieval number prediction result.

[0047] Step 5: Obtain predicted values based on the prediction results. The predicted values include the predicted ovarian response level, the predicted cycle cancellation risk, and the predicted number of retrieved oocytes.

[0048] Specifically, use a link function and based on the prediction results, obtain the predicted values.

[0049] It should be noted that when constructing the dataset based on the clinical characteristic variables in Step 3, it should be determined that the COS patients corresponding to the clinical characteristic variables in the dataset should be patients who exclude any contraindications to assisted reproductive technology and meet the conditions for COS. Clinically, the patients should meet the treatment conditions for receiving classic COS (i.e., agonist protocol and antagonist protocol).

[0050] As an alternative implementation, in order to better construct the prediction model and improve the accuracy of the prediction model, in Step 1, construct a training dataset based on the clinical characteristic variables of historical COS patients, including: dividing the clinical characteristic variables of historical COS patients into continuous variables and categorical variables.

[0051] For example, among the clinical characteristic variables of COS patients: female age (continuous variable), ovulation induction protocol (categorical variable), previous poor ovarian response (categorical variable), previous high ovarian response (categorical variable), infertility duration (continuous variable), thyroid abnormality diagnosis (categorical variable), polycystic ovary syndrome (categorical variable), endometriosis (categorical variable), age at menarche (continuous variable), menstrual period longer than A days (categorical variable), menstrual period shorter than B days (categorical variable), heavy menstrual flow (categorical variable), menstrual cycle shorter than C days (categorical variable), menstrual cycle longer than D days (categorical variable), amenorrhea (categorical variable), irregular menstrual cycle (categorical variable), height (continuous variable), weight (continuous variable), body mass index (continuous variable), basal follicle-stimulating hormone (continuous variable), basal luteinizing hormone (continuous variable), basal prolactin (continuous variable), basal estrogen (continuous variable), basal progesterone (continuous variable), basal testosterone (continuous variable), basal anti-Müllerian hormone (continuous variable), left basal antral follicles (continuous variable), right basal antral follicles (continuous variable), total antral follicle count (continuous variable), starting dose (continuous variable), starting timing (continuous variable), follicle-stimulating hormone on the starting day (continuous variable), luteinizing hormone on the starting day (continuous variable), estrogen on the starting day (continuous variable), AMH on the starting day (continuous variable), human menopausal gonadotropin HMG (continuous variable), recombinant FSH (continuous variable), urinary FSH (continuous variable), follow-up interval (continuous variable), ovarian cysts (categorical variable), mean follicle diameter on the starting day (continuous variable), number of follicles on the starting day (continuous variable), standard deviation of follicle diameter on the starting day (continuous variable).

[0052] Perform dummy variable transformation on the categorical variable to obtain the transformed categorical variable. Fit the continuous variable according to the observed values corresponding to the clinical feature variables to obtain the fitting result. Determine whether there is a linear relationship between the observed values corresponding to the clinical feature variables and the continuous variable based on the fitting result. If there is no linear relationship, transform the continuous variable to obtain the relevant variable; if there is a linear relationship, no processing is performed. Construct an initial training data set based on the transformed categorical variable, the continuous variable, and the relevant variable. Construct a training data set based on the initial training data set.

[0053] The transformation formula for the relevant variable is as follows:

[0054] Basal follicle-stimulating hormone 1 = (MAX(basal follicle-stimulating hormone - 4.71, 0)^3 - ((12.41 - 4.71) * MAX(1.6687921402 - 8.47, 0)^3 + (8.47 - 4.71) * MAX(1.6687921402 - 12.41, 0)^3) / (12.41 - 8.47)) / 59.29.

[0055] Basal follicle-stimulating hormone 2 = (MAX(basal follicle-stimulating hormone - 6.18, 0)^3 - ((12.41 - 6.18) * MAX(1.6687921402 - 8.47, 0)^3 + (8.47 - 6.18) * MAX(1.6687921402 - 12.41, 0)^3) / (12.41 - 8.47)) / 59.29.

[0056] Basal follicle-stimulating hormone 3 = (MAX(basal follicle-stimulating hormone - 7.19, 0)^3 - ((12.41 - 7.19) * MAX(1.6687921402 - 8.47, 0)^3 + (8.47 - 7.19) * MAX(1.6687921402 - 12.41, 0)^3) / (12.41 - 8.47)) / 59.29.

[0057] Basal luteinizing hormone 1 = (MAX(basal luteinizing hormone - 2.03, 0)^3 - ((9.98 - 2.03) * MAX(-0.0316917795 - 5.61, 0)^3 + (5.61 - 2.03) * MAX(-0.0316917795 - 9.98, 0)^3) / (9.98 - 5.61)) / 62.5.

[0058] Basal luteinizing hormone 2 = (MAX(Basal luteinizing hormone - 3.34, 0)^3 - ((9.98 - 3.34)*MAX(-0.0316917795 - 5.61, 0)^3 + (5.61 - 3.34)*MAX(-0.0316917795 - 9.98, 0)^3) / (9.98 - 5.61)) / 62.5.

[0059] Basal luteinizing hormone 3 = (MAX(Basal luteinizing hormone - 4.335, 0)^3 - ((9.98 - 4.335)*MAX(-0.0316917795 - 5.61, 0)^3 + (5.61 - 4.335)*MAX(-0.0316917795 - 9.98, 0)^3) / (9.98 - 5.61)) / 62.5.

[0060] Basal prolactin 1 = (MAX(Basal prolactin - 6.34, 0)^3 - ((31.01 - 6.34)*MAX(.-18.63, 0)^3 + (18.63 - 6.34)*MAX(.-31.01, 0)^3) / (31.01 - 18.63)) / 608.6.

[0061] Basal prolactin 2 = (MAX(Basal prolactin - 10.58725, 0)^3 - ((31.01 - 10.58725)*MAX(.-18.63, 0)^3 + (18.63 - 10.58725)*MAX(-31.01, 0)^3) / (31.01 - 18.63)) / 608.6.

[0062] Basal prolactin 3 = (MAX(Basal prolactin - 13.99, 0)^3 - ((31.01 - 13.99)*MAX(-18.63, 0)^3 + (18.63 - 13.99)*MAX(-31.01, 0)^3) / (31.01 - 18.63)) / 608.6.

[0063] Basal estrogen 1 = (MAX(Basal estrogen - 17, 0)^3 - ((86 - 17)*MAX(-0.5936797091 - 53, 0)^3 + (53 - 17)*MAX(-0.5936797091 - 86, 0)^3) / (86 - 53)) / 4761.

[0064] Basal estrogen 2 = (MAX(Basal estrogen - 31, 0)^3 - ((86 - 31)*MAX(-0.5936797091 - 53, 0)^3 + (53 - 31)*MAX(-0.5936797091 - 86, 0)^3) / (86 - 53)) / 4761.

[0065] Basal estrogen 3 = (MAX(basal estrogen - 41, 0)^3 - ((86 - 41)*MAX(-0.5936797091 - 53, 0)^3 + (53 - 41)*MAX(-0.5936797091 - 86, 0)^3) / (86 - 53)) / 4760.

[0066] Basal progesterone 1 = (MAX(basal progesterone - 0.19, 0)^3 - ((1.7 - 0.19)*MAX(-0.88, 0)^3 + (0.88 - 0.19)*MAX(-1.7, 0)^3) / (1.7 - 0.88)) / 2.28.

[0067] Basal progesterone 2 = (MAX(basal progesterone - 0.41725, 0)^3 - ((1.7 - 0.41725)*MAX(-0.88, 0)^3 + (0.88 - 0.41725)*MAX(-1.7, 0)^3) / (1.7 - 0.88)) / 2.28.

[0068] Basal progesterone 3 = (MAX(basal progesterone - 0.61, 0)^3 - ((1.7 - 0.61)*MAX(-0.88, 0)^3 + (0.88 - 0.61)*MAX(-1.7, 0)^3) / (1.7 - 0.88)) / 2.28.

[0069] Basal testosterone 1 = (MAX(basal testosterone - 0.16, 0)^3 - ((0.92 - 0.16)*MAX(-0.57, 0)^3 + (0.57 - 0.16)*MAX(-0.92, 0)^3) / (0.92 - 0.57)) / 0.58.

[0070] Basal testosterone 2 = (MAX(basal testosterone - 0.33, 0)^3 - ((0.92 - 0.33)*MAX(-0.57, 0)^3 + (0.57 - 0.33)*MAX(-0.92, 0)^3) / (0.92 - 0.57)) / 0.58.

[0071] Basal testosterone 3 = (MAX(basal testosterone - 0.44, 0)^3 - ((0.92 - 0.44)*MAX(-0.57, 0)^3 + (0.57 - 0.44)*MAX(-0.92, 0)^3) / (0.92 - 0.57)) / 0.58.

[0072] Basal anti-Müllerian hormone 1 = (MAX(Basal anti-Müllerian hormone - 0.57, 0)^3 - ((9.0305 - 0.57)*MAX(-0.1524765632 - 4.38, 0)^3 + (4.38 - 0.57)*MAX(-0.1524765632 - 9.0305, 0)^3) / (9.0305 - 4.38)) / 71.6.

[0073] Basal anti-Müllerian hormone 2 = (MAX(Basal anti-Müllerian hormone - 1.67, 0)^3 - ((9.0305 - 1.67)*MAX(-0.1524765632 - 4.38, 0)^3 + (4.38 - 1.67)*MAX(-0.1524765632 - 9.0305, 0)^3) / (9.0305 - 4.38)) / 71.6.

[0074] Basal anti-Müllerian hormone 3 = (MAX(Basal anti-Müllerian hormone - 2.79, 0)^3 - ((9.0305 - 2.79)*MAX(-0.1524765632 - 4.38, 0)^3 + (4.38 - 2.79)*MAX(-0.1524765632 - 9.0305, 0)^3) / (9.0305 - 4.38)) / 71.6.

[0075] Total number of antral follicles 1 = (MAX(Total number of antral follicles - 5, 0)^3 - ((30 - 5)*MAX(-0.1287506639 - 20, 0)^3 + (20 - 5)*MAX(-0.1287506639 - 30, 0)^3) / (30 - 20)) / 625.

[0076] Total number of antral follicles 2 = (MAX(Total number of antral follicles - 10, 0)^3 - ((30 - 10)*MAX(-0.1287506639 - 20, 0)^3 + (20 - 10)*MAX(-0.1287506639 - 30, 0)^3) / (30 - 20)) / 625.

[0077] Total number of antral follicles 3 = (MAX(Total number of antral follicles - 15, 0)^3 - ((30 - 15)*MAX(-0.1287506639 - 20, 0)^3 + (20 - 15)*MAX(-0.1287506639 - 30, 0)^3) / (30 - 20)) / 625.

[0078] Starting quantity 1 = (MAX(Starting quantity - 112.5, 0)^3 - ((-112.5)*MAX(-0.1870746333, 0)^3 + (-112.5)*MAX(-0.1870746333, 0)^3) / (-)) / 12656.25.

[0079] Starting timing 1 = (MAX(Starting timing - 5, 0)^3 - ((30 - 5)*MAX(-20, 0)^3 + (20 - 5)*MAX(-30, 0)^3) / (30 - 20)) / 625.

[0080] Starting timing 2 = (MAX(Starting timing - 10, 0)^3 - ((30 - 10)*MAX(-20, 0)^3 + (20 - 10)*MAX(-30, 0)^3) / (30 - 20)) / 625.

[0081] Starting timing 3 = (MAX(Starting timing - 15, 0)^3 - ((30 - 15)*MAX(-20, 0)^3 + (20 - 15)*MAX(-30, 0)^3) / (30 - 20)) / 625.

[0082] Follicle-stimulating hormone on the starting day 1 = (MAX(Follicle-stimulating hormone on the starting day - 1.04, 0)^3 - ((11.1605 - 1.04)*MAX(0.2334781053 - 4.72, 0)^3 + (4.72 - 1.04)*MAX(0.2334781053 - 11.1605, 0)^3) / (11.1605 - 4.72)) / 102.4.

[0083] Follicle-stimulating hormone on the starting day 2 = (MAX(Follicle-stimulating hormone on the starting day - 1.87, 0)^3 - ((11.1605 - 1.87)*MAX(0.2334781053 - 4.72, 0)^3 + (4.72 - 1.87)*MAX(0.2334781053 - 11.1605, 0)^3) / (11.1605 - 4.72)) / 102.4.

[0084] Follicle-stimulating hormone on the starting day 3 = (MAX(Follicle-stimulating hormone on the starting day - 2.73, 0)^3 - ((11.1605 - 2.73)*MAX(0.2334781053 - 4.72, 0)^3 + (4.72 - 2.73)*MAX(0.2334781053 - 11.1605, 0)^3) / (11.1605 - 4.72)) / 102.4.

[0085] LH on the starting day 1 = (MAX(LH on the starting day - 0.37, 0)^3 - ((5.1 - 0.37) * MAX(0.448238596 - 1.45, 0)^3 + (1.45 - 0.37) * MAX(0.448238596 - 5.1, 0)^3) / (5.1 - 1.45)) / 22.37.

[0086] LH on the starting day 2 = (MAX(LH on the starting day - 0.65, 0)^3 - ((5.1 - 0.65) * MAX(0.448238596 - 1.45, 0)^3 + (1.45 - 0.65) * MAX(0.448238596 - 5.1, 0)^3) / (5.1 - 1.45)) / 22.37.

[0087] LH on the starting day 3 = (MAX(LH on the starting day - 0.89, 0)^3 - ((5.1 - 0.89) * MAX(0.448238596 - 1.45, 0)^3 + (1.45 - 0.89) * MAX(0.448238596 - 5.1, 0)^3) / (5.1 - 1.45)) / 22.37.

[0088] Estradiol on the starting day 1 = (MAX(Estradiol on the starting day - 5, 0)^3 - ((60 - 5) * MAX(.-33, 0)^3 + (33 - 5) * MAX(-60, 0)^3) / (60 - 33)) / 3025.

[0089] Estradiol on the starting day 2 = (MAX(Estradiol on the starting day - 14, 0)^3 - ((60 - 14) * MAX(-33, 0)^3 + (33 - 14) * MAX(.-60, 0)^3) / (60 - 33)) / 3025.

[0090] Estradiol on the starting day 3 = (MAX(Estradiol on the starting day - 22, 0)^3 - ((60 - 22) * MAX(-33, 0)^3 + (33 - 22) * MAX(-60, 0)^3) / (60 - 33)) / 3025.

[0091] Follow - up interval 1 = (MAX(Follow - up interval - 2, 0)^3 - ((6 - 2) * MAX(-0.1359657623 - 5, 0)^3 + (5 - 2) * MAX(-0.1359657623 - 6, 0)^3) / (6 - 5)) / 16.

[0092] Follow-up interval 2 = (MAX(Follow-up interval - 3, 0)^3 - ((6 - 3)*MAX(-0.1359657623 - 5, 0)^3 + (5 - 3)*MAX(-0.1359657623 - 6, 0)^3) / (6 - 5)) / 16.

[0093] Follow-up interval 3 = (MAX(Follow-up interval - 4, 0)^3 - ((6 - 4)*MAX(-0.1359657623 - 5, 0)^3 + (5 - 4)*MAX(-0.1359657623 - 6, 0)^3) / (6 - 5)) / 16.

[0094] Mean follicle diameter on start date 1 = (MAX(Mean follicle diameter on start date - 3.5, 0)^3 - ((6.82 - 3.5)*MAX(-0.3729735097 - 5.133333, 0)^3 + (5.133333 - 3.5)*MAX(-0.3729735097 - 6.82, 0)^3) / (6.82 - 5.133333)) / 11.022.

[0095] Mean follicle diameter on start date 2 = (MAX(Mean follicle diameter on start date - 4.166667, 0)^3 - ((6.82 - 4.166667)*MAX(-0.3729735097 - 5.133333, 0)^3 + (5.133333 - 4.166667)*MAX(-0.3729735097 - 6.82, 0)^3) / (6.82 - 5.133333)) / 11.022.

[0096] Mean follicle diameter on start date 3 = (MAX(Mean follicle diameter on start date - 4.555556, 0)^3 - ((6.82 - 4.555556)*MAX(-0.3729735097 - 5.133333, 0)^3 + (5.133333 - 4.555556)*MAX(-0.3729735097 - 6.82, 0)^3) / (6.82 - 5.133333)) / 11.022.

[0097] Number of follicles on start date 1 = (MAX(Number of follicles on start date - 2, 0)^3 - ((14 - 2)*MAX(0.8401645056 - 9, 0)^3 + (9 - 2)*MAX(0.8401645056 - 14, 0)^3) / (14 - 9)) / 144.

[0098] Number of follicles on start day 2 = (MAX(Number of follicles on start day - 4, 0)^3 - ((14 - 4)*MAX(0.8401645056 - 9, 0)^3 + (9 - 4)*MAX(0.8401645056 - 14, 0)^3) / (14 - 9)) / 144 Number of follicles on start day 3 = (MAX(Number of follicles on start day - 7, 0)^3 - ((14 - 7)*MAX(0.8401645056 - 9, 0)^3 + (9 - 7)*MAX(0.8401645056 - 14, 0)^3) / (14 - 9)) / 144.

[0099] Standard deviation of follicle diameter on start day 1 = (MAX(Standard deviation of follicle diameter on start day - 0, 0)^3 - ((3.8183766 - 0)*MAX(-0.5101505594 - 0.8485281, 0)^3 + (0.8485281 - 0)*MAX(-0.5101505594 - 3.8183766, 0)^3) / (3.8183766 - 0.8485281)) / 14.58.

[0100] Standard deviation of follicle diameter on start day 2 = (MAX(Standard deviation of follicle diameter on start day - 0.212132, 0)^3 - ((3.8183766 - 0.212132)*MAX(-0.5101505594 - 0.8485281, 0)^3 + (0.8485281 - 0.212132)*MAX(-0.5101505594 - 3.8183766, 0)^3) / (3.8183766 - 0.8485281)) / 14.58.

[0101] Standard deviation of follicle diameter on start day 3 = (MAX(Standard deviation of follicle diameter on start day - 0.3535534, 0)^3 - ((3.8183766 - 0.3535534)*MAX(-0.5101505594 - 0.8485281, 0)^3 + (0.8485281 - 0.3535534)*MAX(-0.5101505594 - 3.8183766, 0)^3) / (3.8183766 - 0.8485281)) / 14.58.

[0102] Ratio of FSH to HMG 1 = (MAX(Ratio of FSH to HMG - 0, 0)^3 - ((1 - 0)*MAX(-0.9, 0)^3 + (0.9 - 0)*MAX(-1, 0)^3) / (1 - 0.9)).

[0103] FSH to HMG ratio 2 = (MAX(FSH to HMG ratio - 0.15, 0)^3 - ((1 - 0.15)*MAX(-0.9, 0)^3 + (0.9 - 0.15)*MAX(-1, 0)^3) / (1 - 0.9)).

[0104] FSH to HMG ratio 3 = (MAX(FSH to HMG ratio - 0.375, 0)^3 - ((1 - 0.375)*MAX(-0.9, 0)^3 + (0.9 - 0.375)*MAX(-1, 0)^3) / (1 - 0.9)).

[0105] FSH to HMG ratio 4 = (MAX(FSH to HMG ratio - 0.55, 0)^3 - ((1 - 0.55)*MAX(-0.9, 0)^3 + (0.9 - 0.55)*MAX(-1, 0)^3) / (1 - 0.9)).

[0106] FSH to HMG ratio 5 = (MAX(FSH to HMG ratio - 0.7, 0)^3 - ((1 - 0.7)*MAX(-0.9, 0)^3 + (0.9 - 0.7)*MAX(-1, 0)^3) / (1 - 0.9)).

[0107] The Lasso method is used to perform feature screening on the initial training dataset to obtain the training dataset. Performing feature screening on the initial training dataset includes:

[0108] Based on the Least absolute shrinkage and selection operator (Lasso) method, variables in the initial training dataset are screened. Lasso selects variables by introducing a penalty term into the multiple linear regression model. The training dataset is obtained based on the selected variables. The penalty term is expressed as: λ∑|β j j|. Where ∑ is the Greek symbol representing the sum. The value range of j is from 1 to m, and m is the number of variables in the initial training dataset. β j is the regression coefficient of the jth variable. λ is a constant, and λ ≥ 0, which is used to control the intensity of the penalty term. When λ = 0, all variables are included. When λ increases, the intensity of the penalty term increases, and more variables are excluded and not selected.

[0109] The selection of λ is achieved by generating the lowest possible test Mean Squared Error (MSE). MSE is defined as where n is the total number of observations, that is, n historical COS patient data are obtained. i is a natural number from 1 to n. Y i is the observation value of the ith historical COS patient. yi is the predicted value for the i-th historical COS patient. R software can be used, and the generalized linear model (glmnet) can be employed. By randomly dividing the initial training set into the first initial training set and the second initial training set at a ratio of 9:1, a temporary prediction model is constructed based on the first initial training data set. According to this model and combined with the second initial training set, the temporary predicted value y i , is compared with the observed value Y in the second initial training set i . This process is repeated 10 rounds, and the data used in each round is different, performing 10-fold cross-validation. The test MSE is the average of the MSEs in the 10-fold cross-validation, and then the λ value corresponding to the minimum test MSE is obtained. After substituting the λ value into the model, the regression coefficient of each variable and the intercept of the model can be obtained. Variables with a regression coefficient of 0 are not selected, and thus the training data set is obtained based on the finally obtained regression coefficients, intercepts, and the initial training data set

[0110] The data table of the training data set is shown in Table 1. It should be noted that in all embodiments provided in this application, A is 7, B is 2, C is 21, D is 35; among the clinical characteristic variables, the antagonist protocol is selected for the ovulation induction protocol

[0111] Table 1 Data Table of the Training Data Set

[0112]

[0113]

[0114]

[0115]

[0116] An initial prediction model is constructed based on the training data set. And the initial prediction model is corrected based on the validation data set to obtain the prediction model. The correction process includes:

[0117] The clinical characteristic variables of the historical COS patients in the validation data set are input into the initial prediction model to obtain the initial prediction result. Based on the initial prediction result, the initial predicted value is obtained. The initial predicted value includes: the initial predicted value of the number of retrieved oocytes, the initial predicted value of the cycle cancellation risk, and the initial predicted value of the ovarian response level

[0118] Based on the observed values corresponding to the clinical characteristic variables of historical COS patients in the validation dataset, the initial predicted values of the cycle cancellation risk and the initial predicted values of the ovarian response level are calibrated using a calibration curve, and the initial predicted values of the number of retrieved oocytes are calibrated using the correlation coefficient to obtain the calibrated predicted values. The calibrated predicted values include: the calibrated predicted value of the number of retrieved oocytes, the calibrated predicted value of the cycle cancellation risk, and the calibrated predicted value of the ovarian response level. Based on the calibrated predicted values, the initial prediction model is corrected to obtain the prediction model.

[0119] The data of historical COS patients in the above training dataset and validation dataset are shown in Table 2. It should be noted that in the embodiments provided in this application, the number of historical COS patients in the obtained historical COS patient data is 11,396. Among them, the number of historical COS patients in the training dataset is 9,984, and the number of historical COS patients in the validation dataset is 1,412. In Table 2, Mean represents the average value, SD is the standard deviation, Median represents the median, Min represents the minimum value, and Max represents the maximum value.

[0120] Table 2 Data Table of Historical COS Patients

[0121]

[0122]

[0123]

[0124]

[0125] For example, for the initial prediction models of the cycle cancellation risk and the ovarian response level, the Receiver Operator Characteristic (ROC) curve is used to test the discriminatory power of the initial prediction models in the validation dataset. The ROC curve graphs of the cycle cancellation risk prediction and the ovarian response level prediction are respectively as Figure 3 、 Figure 7 、 Figure 8 and Figure 9 shown. The overall prediction ability of the ovarian response level prediction model is evaluated by macroROC and microROC. The macroROC curve graph and the microROC curve graph are respectively as Figure 5 and Figure 6 shown. The area under the macroROC curve is 0.852 (95% confidence interval of the area under the curve: 0.835 - 0.869); the area under the microROC curve is 0.929 (95% confidence interval of the area under the curve: 0.917 - 0.939).Figure 3 Among them, the area under the ROC curve is 0.897, and the 95% confidence interval of the area under the curve is: 0.8752 - 0.9188. It can be seen that the cycle cancellation risk prediction model has high discriminatory ability. The discriminatory ability of the ovarian response level prediction model is represented by the ROC curve as follows in turn Figure 7 , Figure 8 and Figure 9 shown. The area under their curves is 0.867 (95% confidence interval of the area under the curve: 0.846 - 0.889), 0.797 (95% confidence interval of the area under the curve: 0.775 - 0.824), and 0.893 (95% confidence interval of the area under the curve: 0.866 - 0.917) respectively. It can be seen that the ovarian response level prediction model has high discriminatory ability. The calibration curves are used to test the consistency between the predicted values of the initial prediction model and the corresponding observed values in the validation dataset. The calibration curve graphs of the cycle cancellation risk prediction and the ovarian response level prediction are respectively as Figure 4 , Figure 10 , Figure 11 and Figure 12 shown. The specific method is as follows: taking the predicted value as the abscissa and the observed value as the ordinate, perform linear and smooth fitting on the historical COS patient data of each patient in the validation dataset. According to the results of the linear fitting, obtain the slope and intercept of the fitting curve, and calibrate the predicted value according to the slope and intercept. The calibration method is expressed as: p' = slope * p + intercept. Where p' is the calibrated predicted value, p is the predicted value before calibration, slope is the slope obtained from the above fitting, and intercept is the intercept obtained from the above fitting. The calibrated predicted value is re-smoothed with the observed value, and the judgment criterion for whether the fitting result is accurate is the consistency between the fitting curve and the straight line with a slope of 1 and an intercept of 0 in the coordinate system. Figure 4 Among them, the predicted values of the cycle cancellation risk prediction model are near the ideal curve (predicted value and observed value 1:1), and the prediction is relatively consistent. The calibration curves of the ovarian response level prediction model are as Figure 10 , Figure 11 and Figure 12 shown, indicating that the model may overestimate the possibility of low response and underestimate the possibility of appropriate response and high response in some intervals, but there is no overall estimation of the deviation.

[0126] For the initial prediction model of the number of retrieved oocytes, the correlation coefficient between the predicted value and the observed value is used to represent the prediction consistency. Among them, the scatter plot of the predicted value and the observed value of the number of retrieved oocytes is as Figure 2 shown. The correlation coefficient r value is 0.707, 95% CI (95% confidence interval of the correlation coefficient r): 0.68 - 0.73. P < 0.0001, and P is the statistical significance of the correlation coefficient.

[0127] The finally obtained prediction models include: the oocyte retrieval number prediction model, the cycle cancellation risk prediction model, and the ovarian response level prediction model.

[0128] The general formula of the prediction model is expressed as: y = intercept + β1x1 + … + β j x j . Among them, the value range of j is from 1 to m, and m is the number of variables in the training dataset. In the embodiments provided in this application, m is 95. β j is the regression coefficient of the j-th variable. It should be noted that the number of variables in the initial training dataset is the same as that in the training dataset, so both are represented by m.

[0129] Among them, the oocyte retrieval number prediction model is expressed as: y 获卵数=-1.2401979649+(-0.0147110018)*Female age+(-0.1905561066)*Antagonist protocol+(-0.1949838291)*Previous ovarian hyporesponse+(0.0719372367)*Previous ovarian hyperresponse+(0.0007520638)*Infertility duration+(-0.0314358198)*Thyroid disorder diagnosis+(-0.041043801)*Polycystic ovary syndrome+(-0.0244762497)*Endometriosis+(0.0050370076)*Age at menarche+(-0.0356730278)*Menstrual period longer than A days+(-0.1165025136)*Menstrual period shorter than B days+(0.1130033668)*Heavy menstrual bleeding+(-0.162348083)*Menstrual cycle shorter than C days+(-0.0957179458)*Menstrual cycle longer than D days+(0.2419352851)*Amenorrhea+(-0.1291472585)*Irregular menstrual cycle+(0.0175489881)*Height+(-0.0216725292)*Weight+(0.0465090803)*Body mass index+(-0.0229232895)*Basal follicle-stimulating hormone+(-0.035339131)*Basal follicle-stimulating hormone 1+(0.0026686003)*Basal follicle-stimulating hormone 2+(0.1840211664)*Basal follicle-stimulating hormone 3+(0.024778843)*Basal luteinizing hormone+(-0.0526388808)*Basal luteinizing hormone 2+(-0.0006795541)*Basal prolactin+(0.0132570272)*Basal prolactin 1+(-0.0513082666)*Basal prolactin 3+(0.0002383553)*Basal estrogen+(-0.0004180227)*Basal estrogen 1+(0.0014379944)*Basal estrogen 3+(-0.017837536)*Basal progesterone+(0.0420759255)*Basal progesterone 3+(0.048257276)*Basal testosterone+(-0.0172300434)*Basal testosterone 1+(-0.0240842644)*Basal testosterone 2+(-0.0044862433)*Basal testosterone 3+(0.2436566659)*Basal anti-Müllerian hormone+(-0.6587049355)*Basal anti-Müllerian hormone 1+(0.2514077644)*Basal anti-Müllerian hormone 2+(1.07695189)*Basal anti-Müllerian hormone 3+(0.0024104795)*Left basal antral follicles+(0.0221387699)*Total antral follicle count+(-0.0805824958)*Total antral follicle count 1+(0.1594204074)*Total number of antral follicles 2 + (-0.034449303)*Total number of antral follicles 3 + (0.0024022931)*Starting dose + (-0.0037201176)*Starting dose 1 + (-0.0122049786)*Starting timing + (0.0126152253)*Starting timing 1 + (-0.0180621162)*Follicle-stimulating hormone on the starting day + (0.0273216404)*Follicle-stimulating hormone 1 on the starting day + (-0.12318708)*Follicle-stimulating hormone 3 on the starting day + (0.061354584)*Luteinizing hormone on the starting day + (-0.0688896125)*Luteinizing hormone 1 on the starting day + (-0.0128909446)*Luteinizing hormone 2 on the starting day + (-0.0235390026)*Luteinizing hormone 3 on the starting day + (-0.0004568066)*Estrogen on the starting day + (0.0002293505)*Estrogen 3 on the starting day + (0.1177793809)*Follow-up interval + (-0.1029749343)*Follow-up interval 1 + (-0.2722923527)*Follow-up interval 2 + (1.095161844)*Follow-up interval 3 + (-0.0261389459)*Mean follicle diameter on the starting day + (0.1153319653)*Mean follicle diameter 1 on the starting day + (0.0310351257)*Mean follicle diameter 2 on the starting day + (-0.2188285724)*Mean follicle diameter 3 on the starting day + (0.0345955487)*Number of follicles on the starting day + (0.006091753)*Number of follicles 1 on the starting day + (-0.0650358352)*Number of follicles 2 on the starting day + (0.1359135645)*Number of follicles 3 on the starting day + (-0.1341252662)*Standard deviation of follicle diameter 1 on the starting day + (0.0479053095)*Standard deviation of follicle diameter 3 on the starting day + (0.1079446778)*Ovarian cyst + (-0.1123884471)*Urine-derived FSH + (-0.0689203034)*Ratio of FSH to HMG + (0.1119702599)*Ratio of FSH to HMG 2 + (-4.2540159427)*Ratio of FSH to HMG 5..

[0130] The cycle cancellation risk prediction model is expressed as: y 周期取消风险= 1.6687921402 + (-0.0316917795) * Female age + (-0.5936797091) * Previous ovarian hyporesponse + (-0.1524765632) * Thyroid disorder diagnosis + (-0.1287506639) * Polycystic ovary syndrome + (-0.1870746333) * Endometriosis + (0.2334781053) * Menstrual period longer than A days + (0.448238596) * Menstrual period shorter than B days + (-0.1359657623) * Menstrual cycle shorter than C days + (-0.3729735097) * Menstrual cycle longer than D days + (0.8401645056) * Amenorrhea + (-0.5101505594) * Irregular menstrual cycle + (-0.0164209598) * Body weight + (-0.0723494149) * Body mass index + (-0.1309209177) * Basal follicle-stimulating hormone + (0.0483184945) * Basal luteinizing hormone + (-0.2152163555) * Basal luteinizing hormone 3 + (0.0201272251) * Basal prolactin + (-0.0480120509) * Basal prolactin 3 + (0.0001431656) * Basal estrogen 1 + (0.0226734519) * Basal testosterone + (0.5033537815) * Basal anti-Müllerian hormone + (-0.0095370809) * Basal anti-Müllerian hormone 2 + (-0.9477853116) * Basal anti-Müllerian hormone 3 + (0.0224747932) * Left basal antral follicles + (0.0041659373) * Total antral follicles + (0.2296721781) * Total antral follicles 3 + (-0.0071432889) * Starting dose 1 + (0.1344641971) * Starting timing 3 + (-0.0648560767) * Starting day follicle-stimulating hormone + (-0.0965198036) * Starting day follicle-stimulating hormone 3 + (0.011826389) * Starting day luteinizing hormone + (-0.1312890893) * Follow-up interval + (-0.1724332735) * Mean starting day follicle diameter + (0.0708822652) * Starting day follicle count + (-0.2564592881) * Starting day follicle count 3 + (0.5932646246) * Starting day follicle diameter standard deviation 1 + (0.2149346006) * Starting day follicle diameter standard deviation 3 + (0.0546220283) * Ovarian cyst + (-0.3537510498) * Urinary FSH + (-0.2882333703) * FSH to HMG ratio.

[0131] The ovarian response level prediction model is divided into three types: low response, appropriate response, and high response, which are respectively represented as:

[0132] 1) The response is low (i.e., the predicted number of retrieved oocytes is less than 12).

[0133] y 卵巢响应水平= 4.810244 + (0.05358695) * Female age + (0.5833309) * Previous ovarian hyporesponse + (-0.3817683) * Previous ovarian hyperresponse + (-0.003108974) * Infertility duration + (0.1127085) * Thyroid disorder diagnosis + (0.2156866) * Polycystic ovary syndrome + (0.06725538) * Endometriosis + (-0.02481423) * Age at menarche + (0.02373232) * Menstrual period longer than A days + (2.255483) * Menstrual period shorter than B days + (-0.4137532) * Heavy menstrual bleeding + (0.7448845) * Menstrual cycle shorter than C days + (0.3391707) * Menstrual cycle longer than D days + (-1.158513) * Amenorrhea + (0.4582121) * Irregular menstrual cycle + (-0.01951908) * Height + (0.02522701) * Body mass index + (0.1161208) * Basal follicle-stimulating hormone + (-0.001940659) * Basal follicle-stimulating hormone 1 + (-0.07950017) * Basal luteinizing hormone + (0.3465171) * Basal luteinizing hormone 3 + (-0.01400668) * Basal prolactin + (0.05032311) * Basal prolactin 3 + (-0.0002329462) * Basal estrogen + (-0.0009049723) * Basal estrogen 3 + (0.01477446) * Basal progesterone 3 + (-0.02720292) * Basal testosterone + (0.03313721) * Basal testosterone 3 + (-0.6050613) * Basal anti-Müllerian hormone + (1.053413) * Basal anti-Müllerian hormone 1 + (-2.022607) * Basal anti-Müllerian hormone 3 + (-0.05687471) * Left basal antral follicles + (-0.02833247) * Right basal antral follicles + (-0.005421269) * Total antral follicles + (0.07677861) * Total antral follicles 1 + (-0.5142417) * Total antral follicles 3 + (-0.009810547) * Starting dose + (0.01621158) * Starting dose 1 + (0.005783801) * Starting time + (-0.1155454) * Starting time 3 + (0.03798408) * Follicle-stimulating hormone on starting day + (0.09127777) * Follicle-stimulating hormone 3 on starting day + (-0.02997599) * Luteinizing hormone on starting day + (0.007712081) * Luteinizing hormone 2 on starting day + (0.06626255) * Luteinizing hormone 3 on starting day + (0.001581909) * Estrogen on starting day + (0.00001549964) * Estrogen 2 on starting day + (0.004725451) * Estrogen 3 on starting day + (0.(0.2337314)*Follow-up interval + (0.09694711)*Follow-up interval 1 + (-0.3773947)*Follow-up interval 3 + (0.1285488)*Mean follicle diameter on start date + (0.1384914)*Mean follicle diameter on start date 2 + (0.05889516)*Mean follicle diameter on start date 3 + (-0.07061949)*Number of follicles on start date + (-0.05436386)*Number of follicles on start date 1 + (0.3270961)*Number of follicles on start date 3 + (-0.02693619)*Standard deviation of follicle diameter on start date + (0.4697538)*Standard deviation of follicle diameter on start date 1 + (-0.1101547)*Standard deviation of follicle diameter on start date 3 + (-0.1940838)*Ovarian cyst + (0.6742072)*Urine-derived FSH + (0.4242473)*Ratio of FSH to HMG + (0.04233529)*Ratio of FSH to HMG 3 + (0.2907345)*Ratio of FSH to HMG 4 + (1.264607)*Ratio of FSH to HMG 5.

[0134] 2) Respond appropriately (i.e., the predicted number of retrieved oocytes is between 12 and 18).

[0135] y 卵巢响应水平=-0.6505132+(-0.002333914)*Female age+(-0.168155)*Previous ovarian hyporesponse+(0.2671149)*Previous ovarian hyperresponse+(-0.008340805)*Infertility duration+(-0.216962)*Thyroid disorder diagnosis+(-0.02406286)*Polycystic ovary syndrome+(0.07420737)*Endometriosis+(-0.03380935)*Age at menarche+(0.03272799)*Menstrual period longer than A days+(-1.611643)*Menstrual period shorter than B days+(-0.09984485)*Heavy menstrual flow+(-0.8775878)*Menstrual cycle shorter than C days+(-0.2221219)*Menstrual cycle longer than D days+(0.2721955)*Amenorrhea+(-0.2419314)*Irregular menstrual cycle+(-0.003666913)*Height+(0.004090348)*Body mass index+(-0.00002777771)*Basal follicle-stimulating hormone+(0.02255724)*Basal follicle-stimulating hormone 1+(-0.001557238)*Basal luteinizing hormone+(0.06000859)*Basal luteinizing hormone 3+(-0.001573342)*Basal prolactin+(0.01726146)*Basal prolactin 3+(-0.000004433223)*Basal estrogen+(0.0001381945)*Basal estrogen 3+(0.00272952)*Basal progesterone 3+(0.01017567)*Basal testosterone+(-0.01774593)*Basal testosterone 3+(0.3257581)*Basal anti-Müllerian hormone+(-0.9683422)*Basal anti-Müllerian hormone 1+(2.318202)*Basal anti-Müllerian hormone 3+(0.01436576)*Left basal antral follicles+(0.02865711)*Right basal antral follicles+(0.00343396)*Total antral follicles+(-0.07885612)*Total antral follicles 1+(0.5633314)*Total antral follicles 3+(0.006856208)*Starting dose+(-0.008869535)*Starting dose 1+(-0.0009426046)*Starting time+(0.01183184)*Starting time 3+(-0.02200009)*Follicle-stimulating hormone on starting day+(-0.04260421)*Follicle-stimulating hormone 3 on starting day+(0.02092761)*Luteinizing hormone on starting day+(-0.01089947)*Luteinizing hormone 2 on starting day+(-0.08698512)*Luteinizing hormone 3 on starting day+(-0.0008569579)*Estrogen on starting day+(-0.00002277761)*Estrogen 2 on starting day+(-0.009833575)*Estrogen on start date 3 + (-0.02859406)*Follow-up interval + (-0.04442596)*Follow-up interval 1 + (0.657524)*Follow-up interval 3 + (0.044548)*Mean follicle diameter on start date + (-0.1728033)*Mean follicle diameter on start date 2 + (-0.06669016)*Mean follicle diameter on start date 3 + (0.01879822)*Number of follicles on start date + (0.001595511)*Number of follicles on start date 1 + (-0.05401126)*Number of follicles on start date 3 + (0.01363645)*Standard deviation of follicle diameter on start date + (-0.460303)*Standard deviation of follicle diameter on start date 1 + (0.0546288)*Standard deviation of follicle diameter on start date 3 + (0.359914)*Ovarian cyst + (-0.02787798)*Urinary FSH + (0.003772108)*Ratio of FSH to HMG + (0.09538819)*Ratio of FSH to HMG 3 + (0.4885847)*Ratio of FSH to HMG 4 + (1.181186)*Ratio of FSH to HM 5.

[0136] 3) Excessive response (i.e., predicted number of retrieved oocytes is greater than 18).

[0137] y 卵巢响应水平=-4.159731+(-0.05125303)*Female age+(-0.415176)*Previous ovarian hyporesponse+(0.1146534)*Previous ovarian hyperresponse+(0.01144978)*Infertility duration+(0.1042535)*Thyroid disorder diagnosis+(-0.1916238)*Polycystic ovary syndrome+(-0.1414628)*Endometriosis+(0.05862358)*Age at menarche+(-0.05646031)*Menstrual period longer than A days+(-0.64384)*Menstrual period shorter than B days+(0.513598)*Heavy menstrual flow+(0.1327032)*Menstrual cycle shorter than C days+(-0.1170487)*Menstrual cycle longer than D days+(0.8863179)*Amenorrhea+(-0.2162807)*Irregular menstrual cycle+(0.02318599)*Height+(-0.02931736)*Body mass index+(-0.116093)*Basal follicle-stimulating hormone+(-0.02061658)*Basal follicle-stimulating hormone 1+(0.0810574)*Basal luteinizing hormone+(-0.4065257)*Basal luteinizing hormone 3+(0.01558002)*Basal prolactin+(-0.06758457)*Basal prolactin 3+(0.0002373794)*Basal estrogen+(0.0007667778)*Basal estrogen 3+(-0.01750398)*Basal progesterone 3+(0.01702725)*Basal testosterone+(-0.01539128)*Basal testosterone 3+(0.2793033)*Basal anti-Müllerian hormone+(-0.08507105)*Basal anti-Müllerian hormone 1+(-0.2955952)*Basal anti-Müllerian hormone 3+(0.04250895)*Left basal antral follicles+(-0.0003246412)*Right basal antral follicles+(0.001987309)*Total antral follicles+(0.002077511)*Total antral follicles 1+(-0.04908968)*Total antral follicles 3+(0.002954339)*Starting dose+(-0.007342041)*Starting dose 1+(-0.004841196)*Starting time+(0.1037136)*Starting time 3+(-0.01598399)*Follicle-stimulating hormone on starting day+(-0.04867355)*Follicle-stimulating hormone 3 on starting day+(0.009048378)*Luteinizing hormone on starting day+(0.003187388)*Luteinizing hormone 2 on starting day+(0.02072257)*Luteinizing hormone 3 on starting day+(-0.000724951)*Estrogen on starting day+(0.000007277973)*Estrogen 2 on starting day+(0.005108124)*Estrogen 3 on starting day+(0.005220922)*Follow-up interval + (-0.05252115)*Follow-up interval 1 + (-0.2801293)*Follow-up interval 3 + (-0.1730968)*Mean follicle diameter on the start date + (0.03431195)*Mean follicle diameter on the start date 2 + (0.007794999)*Mean follicle diameter on the start date 3 + (0.05182127)*Number of follicles on the start date + (0.05276835)*Number of follicles on the start date 1 + (-0.2730849)*Number of follicles on the start date 3 + (0.01329974)*Standard deviation of follicle diameter on the start date + (-0.009450845)*Standard deviation of follicle diameter on the start date 1 + (0.0555259)*Standard deviation of follicle diameter on the start date 3 + (-0.1658301)*Ovarian cyst + (-0.6463293)*Urine-derived FSH + (-0.4280195)*Ratio of FSH to HMG + (-0.1377235)*Ratio of FSH to HMG 3 + (-0.7793192)*Ratio of FSH to HMG 4 + (-2.445793)*Ratio of FSH to HMG 5.

[0138] It should be noted that in the above prediction model representation, the variables not shown are the variables with a regression coefficient of 0.

[0139] In addition, in step 5, a link function is used to obtain a predicted value based on the prediction result. Among them, according to the type of the prediction model, different link functions are used to convert the obtained y value into a predicted value.

[0140] For the cycle cancellation risk prediction model and the ovarian response level prediction model, the link function is expressed as: By combining the y values obtained from the cycle cancellation risk prediction model and the ovarian response level prediction model with the link function respectively, the predicted value p corresponding to the y value can be obtained.

[0141] For the oocyte retrieval number prediction model, the link function is expressed as: p = exp(y). By combining the y value obtained from the oocyte retrieval number prediction model with the link function, the predicted value p corresponding to the y value can be obtained.

[0142] Based on the same inventive concept, the embodiments of the present application also provide an ovarian response level and cycle cancellation risk prediction system for implementing the above-mentioned ovarian response level and cycle cancellation risk prediction methods. The implementation solutions provided by this system to solve problems are similar to the implementation solutions described in the above methods. Therefore, the specific limitations in the embodiments of the ovarian response level and cycle cancellation risk prediction system provided below can refer to the limitations on the ovarian response level and cycle cancellation risk prediction methods in the above text, and will not be elaborated here.

[0143] In an exemplary embodiment, an ovarian response level and cycle cancellation risk prediction system is provided, including:

[0144] A data acquisition module, configured to acquire clinical characteristic variables of COS patients.

[0145] A data processing module, connected to the data acquisition module, configured to obtain a feature data set based on the clinical characteristic variables.

[0146] A data prediction module, connected to the data processing module, configured to obtain a prediction result based on the feature data set by using a prediction model, and obtain a predicted value based on the prediction result.

[0147] A display module, connected to the data prediction module, configured to display the predicted value.

[0148] In an exemplary embodiment, a computer device is provided. The computer device can be a server or a terminal, and its internal structure diagram can be as shown in Figure 14 shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store relevant data involved in the ovarian response level and cycle cancellation risk prediction method. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, it implements an ovarian response level and cycle cancellation risk prediction method.

[0149] Those skilled in the art can understand that Figure 14 the structure shown in is only a block diagram of some structures related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps in the above method embodiments are implemented.

[0150] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, which when executed by a processor, implements the steps in the above method embodiments.

[0151] In an exemplary embodiment, a computer program product is provided, including a computer program, which when executed by a processor, implements the steps in the above method embodiments.

[0152] It should be noted that the patient information (including but not limited to patient device information, patient personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the patient personally or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with relevant regulations.

[0153] Those of ordinary skill in the art can understand that all or part of the processes in the above method embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the above method embodiments. Among them, any reference to a memory, database, or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memories. Non-volatile memories 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), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memories can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0154] In each of the embodiments provided in the present application, the database involved may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., without limitation. In each of the embodiments provided in the present application, the processor involved may be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., without limitation.

[0155] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, 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, it should be considered as the scope described in this specification.

[0156] Specific examples are used in this article to elaborate on the principles and implementation manners of the present application. The description of the above embodiments is only used to help understand the method and its core idea of the present application; at the same time, for those of ordinary skill in the art, according to the idea of the present application, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to the present application.

Claims

1. A method for predicting ovarian response level and cycle cancellation risk, characterized in that: The method for predicting ovarian response level and cycle cancellation risk includes: Constructing a prediction model; the prediction model includes: a prediction model for the number of eggs obtained, a prediction model for the risk of cycle cancellation, and a prediction model for the level of ovarian response; Obtain clinical characteristic variables of COS patients; the clinical characteristic variables include: female age, ovulation induction regimen, previous ovarian low response, previous ovarian high response, infertility years, thyroid abnormality diagnosis, polycystic ovary syndrome, endometriosis, age of menarche, menstrual period longer than A day, menstrual period shorter than B day, heavy menstrual flow, menstrual cycle shorter than C day, menstrual cycle longer than D day, amenorrhea, irregular menstrual cycle, height, weight, body mass index, basal follicle stimulating hormone, basal luteinizing hormone, basal prolactin, basal estrogen, basal progesterone, basal testosterone, basal anti-Mullerian hormone, left basal antral follicles, right basal antral follicles, total number of antral follicles, activation amount, activation time, follicle stimulating hormone on the activation day, luteinizing hormone on the activation day, estrogen on the activation day, AMH on the activation day, human menopausal gonadotropin HMG, recombinant FSH, urine-derived FSH, follow-up interval, ovarian cysts, mean follicle diameter on the activation day, number of follicles on the activation day, and standard deviation of follicle diameter on the activation day; constructing a feature data set based on the clinical feature variables; Inputting the characteristic data set into the prediction model to obtain prediction results; the prediction results include ovarian response level prediction results, cycle cancellation risk prediction results and egg number prediction results; A prediction value is obtained based on the prediction result; the prediction value includes an ovarian response level prediction value, a cycle cancellation risk prediction value and an egg number prediction value.

2. The method for predicting ovarian response level and cycle cancellation risk according to claim 1, characterized in that: Building a predictive model involves: Obtaining historical COS patient data with different exogenous Gn usage levels; the historical COS patient data includes clinical characteristic variables of historical COS patients and observation values ​​corresponding to the clinical characteristic variables; Constructing a training data set based on the clinical characteristic variables of the historical COS patients; Constructing a validation dataset based on the historical COS patient data; Building an initial prediction model based on the training data set; The initial prediction model is corrected based on the verification data set to obtain the prediction model.

3. The method for predicting ovarian response level and cycle cancellation risk according to claim 2, characterized in that: Correcting the initial prediction model based on the validation data set to obtain the prediction model includes: Inputting the clinical characteristic variables of historical COS patients in the validation data set into the initial prediction model to obtain an initial prediction result; Based on the initial prediction result, an initial prediction value is obtained; the initial prediction value includes: an initial prediction value of the number of eggs obtained, an initial prediction value of the risk of cycle cancellation, and an initial prediction value of the ovarian response level; Based on the observed values ​​corresponding to the clinical characteristic variables of the historical COS patients in the validation data set, the initial predicted value of the cycle cancellation risk and the initial predicted value of the ovarian response level are calibrated using a calibration curve, and the initial predicted value of the number of oocytes obtained is calibrated using a correlation coefficient to obtain a calibrated predicted value; the calibrated predicted value includes: a calibrated predicted value of the number of oocytes obtained, a calibrated predicted value of the cycle cancellation risk, and a calibrated predicted value of the ovarian response level; Based on the calibrated prediction value, the initial prediction model is corrected to obtain the prediction model.

4. The method for predicting ovarian response level and cycle cancellation risk according to claim 2, characterized in that: A training data set was constructed based on the clinical characteristic variables of the historical COS patients, including: The clinical characteristic variables of the historical COS patients were divided into continuous variables and categorical variables; Performing dummy variable conversion on the categorical variable to obtain a converted categorical variable; Fitting the continuous variable according to the observed value corresponding to the clinical characteristic variable to obtain a fitting result; Determine whether there is a linear relationship between the observed value corresponding to the clinical characteristic variable and the continuous variable based on the fitting result; if there is no linear relationship, transform the continuous variable to obtain a related variable; if there is a linear relationship, do not process; constructing an initial training data set based on the transformed categorical variables, the continuous variables and the correlated variables; A training data set is constructed based on the initial training data set.

5. The method for predicting ovarian response level and cycle cancellation risk according to claim 4, characterized in that: The Lasso method is used to perform feature screening on the initial training data set to obtain the training data set.

6. The method for predicting ovarian response level and cycle cancellation risk according to claim 1, characterized in that: Obtaining a predicted value based on the prediction result includes: The predicted value is obtained based on the predicted result using a link function.

7. A system for predicting ovarian response level and cycle cancellation risk, characterized in that: The ovarian response level and cycle cancellation risk prediction system is used to implement the ovarian response level and cycle cancellation risk prediction method according to any one of claims 1 to 6; The ovarian response level and cycle cancellation risk prediction system includes: Data acquisition module, used to obtain clinical characteristic variables of COS patients; A data processing module, connected to the data acquisition module, for obtaining a characteristic data set based on the clinical characteristic variables; A data prediction module, connected to the data processing module, for obtaining a prediction result based on the feature data set by using a prediction model, and obtaining a prediction value based on the prediction result; A display module is connected to the data prediction module and is used to display the predicted value.

8. A computer device comprising: 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 method for predicting ovarian response level and cycle cancellation risk according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for predicting ovarian response level and cycle cancellation risk according to any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for predicting ovarian response level and cycle cancellation risk according to any one of claims 1 to 6 is implemented.

Images