Embryo transplantation outcome prediction method, device, equipment, medium and product

Abstract

The invention discloses an embryo transplantation outcome prediction method, device, equipment, medium and product, and relates to the field of embryo transplantation, the method comprises the following steps: obtaining historical transplantation data of an embryo transplantation patient; screening and dividing the historical transplantation data to obtain a data set; the data set comprises a training set and an external verification set; performing feature screening on the data set to obtain risk factors; modeling by using the risk factors of the training set and the historical embryo transplantation outcome of the training set to obtain a prediction model; adjusting model parameters of the prediction model by using the external verification set to obtain an adjusted prediction model; and performing prediction by using the adjusted prediction model according to clinical characteristic variables of a pre-embryo transplantation patient to obtain an embryo transplantation outcome. The prediction accuracy of the embryo transplantation outcome is improved.

Description

Technical Field

[0001] This application relates to the field of embryo transfer, and particularly to a method, device, equipment, medium and product for predicting the outcome of embryo transfer. Background Art

[0002] The number of embryo transfers (ET) is one of the most important decisions in the assisted reproductive technologies (ART) process. Transferring multiple embryos simultaneously and having them all implant will result in multiple pregnancies and cause a series of pregnancy risks and complications. Therefore, professional organizations such as the European Society of Human Reproduction and Embryology usually recommend transferring only one embryo (Hum Reprod. 2024 Apr 3; 39(4):647 - 657.). However, the implantation chance of embryos varies according to the patient's clinical condition, the developmental stage of the embryo, and its morphology. Clinically, due to reasons such as the patient's concern about the chance of success, it is difficult to completely avoid transferring multiple embryos. Therefore, predicting the outcomes of different embryo transfer strategies based on the specific morphology, developmental stage of the transferred embryos, and the patient's individualized clinical parameters will help doctors and patients understand the potential risks. Summary of the Invention

[0003] The purpose of this application is to provide a method, device, equipment, medium and product for predicting the outcome of embryo transfer, which can improve the prediction accuracy of the embryo transfer outcome.

[0004] To achieve the above purpose, this application provides the following solutions:

[0005] In the first aspect, this application provides a method for predicting the outcome of embryo transfer, including:

[0006] Obtaining the historical transfer data of embryo transfer patients; the historical transfer data includes historical clinical characteristic variables and historical embryo transfer outcomes; the historical clinical characteristic variables include continuous variables and categorical variables; the historical embryo transfer outcomes include failed transfer, singleton pregnancy or multiple pregnancy;

[0007] Screening and dividing the historical transfer data to obtain a data set; the data set includes a training set and an external validation set;

[0008] Performing feature screening on the data set to obtain risk factors;

[0009] Using the risk factors of the training set and the historical embryo transfer outcomes of the training set to build a prediction model;

[0010] Using the external validation set to adjust the model parameters of the prediction model to obtain an adjusted prediction model;

[0011] Predict using the adjusted prediction model according to the clinical characteristic variables of pre-embryo transfer patients to obtain the embryo transfer outcome.

[0012] In a second aspect, the present application provides an embryo transfer outcome prediction device, including:

[0013] An acquisition module for acquiring historical transplantation data of embryo transfer patients; the historical transplantation data includes historical clinical characteristic variables and historical embryo transfer outcomes; the historical clinical characteristic variables include continuous variables and categorical variables; the historical embryo transfer outcomes include transplantation failure, singleton pregnancy, or multiple pregnancy;

[0014] A screening and partitioning module for screening and partitioning the historical transplantation data to obtain a data set; the data set includes a training set and an external validation set;

[0015] A feature screening module for screening features of the data set to obtain risk factors;

[0016] A modeling module for modeling using the risk factors of the training set and the historical embryo transfer outcomes of the training set to obtain a prediction model;

[0017] An adjustment module for adjusting the model parameters of the prediction model using the external validation set to obtain an adjusted prediction model;

[0018] A prediction module for predicting using the adjusted prediction model according to the clinical characteristic variables of pre-embryo transfer patients to obtain the embryo transfer outcome.

[0019] 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, where the processor executes the computer program to implement the embryo transfer outcome prediction method described in any one of the above.

[0020] 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, it implements the embryo transfer outcome prediction method described in any one of the above.

[0021] 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, it implements the embryo transfer outcome prediction method described in any one of the above.

[0022] According to the specific embodiments provided by the present application, the present application has the following technical effects:

[0023] The present application provides a method, device, equipment, medium and product for predicting the outcome of embryo transfer. By screening and dividing historical transplantation data, a training set and an external validation set are obtained, and feature screening is performed on the data sets to improve the effectiveness of the data sets. Then, a model is constructed using the training set, and the prediction model is adjusted using the external validation set, so that the model is more adapted to the historical transplantation data, and further improve the accuracy of predicting the outcome of embryo transfer using the adjusted prediction model. Description of the Drawings

[0024] 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 to be used in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0025] Figure 1 It is a schematic diagram of the technical modules of a method for predicting the outcome of embryo transfer in an embodiment of the present application;

[0026] Figure 2 It is a macro ROC curve graph of the training set data in an embodiment of the present application;

[0027] Figure 3 It is a micro ROC curve graph of the training set data in an embodiment of the present application;

[0028] Figure 4 It is a macro ROC curve graph in the validation data in an embodiment of the present application;

[0029] Figure 5 It is a micro ROC curve graph in the validation data in an embodiment of the present application

[0030] Figure 6 It is a comparison graph of the areas under the ROC curves of the present model and the existing model in an embodiment of the present application;

[0031] Figure 7 It is a comparison graph of the decision curves of the present model and the existing model;

[0032] Figure 8 It is an application environment diagram of a method for predicting the outcome of embryo transfer in an embodiment of the present application;

[0033] Figure 9 It is a schematic flowchart of a method for predicting the outcome of embryo transfer in an embodiment of the present application;

[0034] Figure 10 It is a schematic structural diagram of a computer device provided in an embodiment of the present application. Detailed Embodiments

[0035] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to 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. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts shall fall within the protection scope of the present application.

[0036] 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 with reference to the accompanying drawings and specific embodiments.

[0037] The embryo transfer outcome prediction method provided by the embodiments of the present application can be applied to an application environment as Figure 8 shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, placed in the cloud or on other servers. The terminal 102 can send the clinical characteristic variables of the pre-embryo transfer patient to the server 104. After receiving the clinical characteristic variables of the pre-embryo transfer patient, for the clinical characteristic variables of the pre-embryo transfer patient, the server 104 screens and divides the historical transplantation data of the embryo transfer patient to obtain a data set; the data set includes a training set and an external validation set; feature screening is performed on the data set to obtain risk factors; a prediction model is obtained by modeling the risk factors of the training set and the historical embryo transfer outcomes of the training set; the model parameters of the prediction model are adjusted using the external validation set to obtain an adjusted prediction model; the embryo transfer outcome is predicted using the adjusted prediction model according to the clinical characteristic variables of the pre-embryo transfer patient. The server 104 can feedback the obtained embryo transfer outcome to the terminal 102. In addition, in some embodiments, the embryo transfer outcome prediction method can also be implemented separately by the server 104 or the terminal 102. For example, the terminal 102 can directly predict the clinical characteristic variables of the pre-embryo transfer patient, or the server 104 can obtain the clinical characteristic variables of the pre-embryo transfer patient from the data storage system and predict the clinical characteristic variables of the pre-embryo transfer patient.

[0038] Among them, the terminal 102 can be, but is not limited to, various desktop computers, laptop computers, smart phones, tablet computers, Internet of Things devices, and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, etc. The portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. The server 104 can be implemented by an independent server or a server cluster composed of multiple servers, and can also be a cloud server.

[0039] In an exemplary embodiment, as Figure 9 shown, a method for predicting the outcome of embryo transfer is provided. This method is executed by a computer device, and specifically can be executed alone by a computer device such as a terminal or a server, or can be jointly executed by a terminal and a server. In the embodiments of the present application, taking this method applied to Figure 8 the server 104 in it as an example for illustration, it includes the following steps, where:

[0040] Step 201: Obtain the historical transplantation data of embryo transfer patients; the historical transplantation data includes historical clinical characteristic variables and historical embryo transfer outcomes; the historical clinical characteristic variables include continuous variables and categorical variables; the historical embryo transfer outcomes include transplantation failure, singleton pregnancy, or multiple pregnancy.

[0041] Step 202: Screen and divide the historical transplantation data to obtain a data set; the data set includes a training set and an external validation set.

[0042] Step 203: Perform feature screening on the data set to obtain risk factors.

[0043] Step 204: Use the risk factors of the training set and the historical embryo transfer outcomes of the training set to build a prediction model.

[0044] Step 205: Use the external validation set to adjust the model parameters of the prediction model to obtain an adjusted prediction model.

[0045] Step 206: Use the adjusted prediction model to predict according to the clinical characteristic variables of pre-embryo transfer patients to obtain the embryo transfer outcome.

[0046] Implementing the above steps 201 to 206, building a model using the training set, and then using the external validation set to adjust the prediction model, thereby improving the prediction accuracy of the adjusted prediction model.

[0047] In an exemplary embodiment, screening and dividing the historical transplantation data to obtain a data set specifically includes:

[0048] Screen the historical transplantation data according to the screening criteria to obtain the screened data; the screening criteria are to exclude the contraindications of assisted reproductive technology and conform to the historical transplantation data of patients undergoing embryo transplantation.

[0049] Perform dummy variable transformation on the categorical variables in the screened data to obtain the transformed categorical variables. In practical applications, the categorical variables include whether there has been a previous live birth, whether there has been a previous spontaneous abortion, endometriosis, polycystic ovary syndrome, uterine morphological abnormalities, uterine fibroids, intrauterine adhesions, cesarean section history, endometrial scar diverticulum, intracytoplasmic sperm injection, whole embryo cryopreservation, transplantation cycle type, fresh down-regulation cycle, fresh non-down-regulation cycle, frozen-thawed down-regulation cycle, frozen-thawed hormone replacement cycle, frozen-thawed natural cycle, frozen-thawed ovulation induction cycle, poor endometrial echo morphology, transplantation strategy, number of transplantations, and morphology of transplanted embryos.

[0050] Use the generalized additive model to fit the continuous variables in the screened data with the historical embryo transplantation outcomes to obtain variables with non-linear relationships, and perform restricted cubic spline transformation on the variables with non-linear relationships to obtain the transformed continuous variables. In practical applications, the continuous variables include female age, female height, female weight, female body mass index, basal follicle-stimulating hormone, basal luteinizing hormone, basal prolactin, basal estrogen, basal progesterone, basal testosterone, basal anti-Müllerian hormone, basal antral follicle count, male age, male height, male weight, male body mass index, infertility duration, starting dose, number of retrieved oocytes, number of mature oocytes, mature oocyte rate, number of fertilized eggs, fertilization rate, number of 2PN fertilized eggs, 2PN fertilization rate, number of cleaved embryos, number of 2PN cleaved embryos, cleavage rate, number of MI oocytes, number of GV oocytes, number of degenerated oocytes, number of embryos, number of high-quality embryos, endometrial thickness, and distance from transplantation to the fundus of the uterus.

[0051] Partition the transformed categorical variables and the transformed continuous variables to obtain a data set.

[0052] In an exemplary embodiment, perform feature screening on the data set to obtain risk factors, specifically including: using the least absolute shrinkage and selection operator method to perform feature screening on the data set to obtain risk factors.

[0053] In an exemplary embodiment, the expression of the prediction model is:

[0054] y = β1X1 + … + βiX i + intercept;

[0055] p = exp(y) / (1 - exp(y));

[0056] where y is the response value, βi is the coefficient of the i-th variable, p is the predicted value, X1 is the first relevant variable, and X i is the i-th relevant variable.

[0057] In another exemplary embodiment of the present application, in order to prove the effectiveness of an embryo transfer outcome prediction method provided by the present application, an application process of an embryo transfer outcome prediction method in practice is also provided. In the assisted reproductive technology process, according to the patient's situation on the day of embryo transfer, the risks of embryo implantation failure and multiple pregnancy, as well as the chance of achieving a singleton pregnancy, are predicted under different ET decisions. ET decision refers to: the morphological score of the embryo to be transferred (according to the Istanbul Consensus, Hum Reprod. 2011 Jun; 26(6): 1270-83), the developmental stage of the embryo (cleavage stage or blastocyst stage), and the number of embryos transferred (1 or 2). The specific steps are as follows:

[0058] S1. Collect the clinical characteristic variables of ET patients, and conduct preliminary data screening and filtering to confirm the inclusion criteria for the selected subjects, and construct a data set. The data set is divided into a training set (development set) and an external validation set (external validation set). The differentiation method can be factors such as the time and location of data collection.

[0059] S2. Use the Least absolute shrinkage and selection operator (Lasso) method to screen the features of the data set, and use the finally screened indicators as risk factors;

[0060] S3. Use the indicators finally screened in step S2 to build a model and conduct performance evaluation;

[0061] S4. Evaluate and calibrate the prediction effect of the model through the external validation set, and readjust the model through the slope and intercept of the calibration curve;

[0062] S5. Build online and offline tools, and use the S3 and S4 models for prediction and interpretation.

[0063] Detailed technical parameters can also obtain the predicted value through manual calculation, office software such as excel, or other programming methods.

[0064] In practical applications, the clinical characteristic variables of patients include:

[0065] Female age (continuous variable), female height (continuous variable), female weight (continuous variable), female 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), basal antral follicle count (continuous variable), male age (continuous variable), male height (continuous variable), male weight (continuous variable), male body mass index (continuous variable), infertility duration (continuous variable), whether there was a previous live birth (yes / no), whether there was a previous spontaneous abortion (none, once, twice, more than three times), endometriosis (yes / no), polycystic ovary syndrome (yes / no), uterine morphological abnormality (yes / no), uterine fibroids (yes / no), intrauterine adhesions (yes / no), cesarean section history (yes / no), endometrial scar diverticulum (yes / no), intracytoplasmic sperm injection (yes / no), starting dose (continuous variable), number of retrieved oocytes (continuous variable), number of mature oocytes (continuous variable), mature oocyte rate (continuous variable), number of fertilized eggs (continuous variable), fertilization rate (continuous variable), number of 2PN fertilized eggs (continuous variable), 2PN fertilization rate (continuous variable), number of cleaved embryos (continuous variable), number of 2PN cleaved embryos (continuous variable), cleavage rate (continuous variable), number of MI oocytes (continuous variable), number of GV oocytes (continuous variable), number of degenerated oocytes (continuous variable), number of embryos (continuous variable), number of high-quality embryos (continuous variable), whole embryo cryopreservation (yes / no), type of transplantation cycle, fresh down-regulation cycle, fresh non-down-regulation cycle (yes / no), frozen-thawed down-regulation cycle (yes / no), frozen-thawed hormone replacement cycle (yes / no), frozen-thawed natural cycle (yes / no), frozen-thawed induced ovulation cycle (yes / no), endometrial thickness (continuous variable), poor endometrial echo morphology (yes / no), distance from transplantation to the uterine fundus (continuous variable), transplantation strategy (single blastocyst transplantation, double blastocyst transplantation, single blastomere transplantation, double blastomere transplantation), number of transplantations (1, 2, ≥3), morphology of transplanted embryos (normal, normal + poor, high-quality, high-quality + normal, high-quality + poor, poor).

[0066] For the indicators that are categorical variables, dummy variable transformation is performed. For those that are continuous variables, fitting with the outcome is carried out through the generalized additive model (GAM) to determine whether there is a linear relationship. For variables with a non-linear relationship, transformation is performed through restricted cubic spline (rcs). Besides the original feature variable (X), k related variables are added, denoted as X1, X2... X k, the original variable and the related variables are jointly incorporated into the model. In the rcs transformation, the number of related variables is from 1 to k, and the number of related variables among the features of the entire model is from 1 to i. There are respective transformation formulas for the 1 - K variables, and 3 variables are the most common case, so there are usually 3 formulas. These variables represent the fitting after splitting the original variable and have no specific physical meaning. Here, it is used as a statistical technique.

[0067] The general formula is that for any continuous variable X, nodes t1 < t2 < ··· < t are set within its value range. n , n = k + 2.

[0068] n is the number of nodes, t is the set node, and t n is the nth node, that is, the last node. t K is the Kth node. 1 <= K < k means that k related variables are set before, and the calculation of any one of the variables is shown as follows. For integers 1 <= K <= k, there is:

[0069] X K == (MAX(X - t k , 0)^3 - ((t n - t K ) * (MAX(X - t n-1 , 0)^3) + (t n-1 - t K ) * (MAX(X - t n , 0)^3)) / (t

[0070] n - t n-1 )) / C.

[0071] C is a constant generated by numerical fitting.

[0072] When n = 2, K = 0, the fitting is linear and no additional variables are generated.

[0073] The transformation formulas of the related variables are as follows.

[0074] Basic antral follicle count 1 = (MAX(Basic antral follicle count - 3, 0)^3 - ((21 - 3) * (MAX(Basic antral follicle count - 13, 0)^3) + (13 - 3) * (MAX(Basic antral follicle count - 21, 0)^3)) / (21 - 13)) / 324.

[0075] Basic antral follicle count 2 = (MAX(Basic antral follicle count - 7, 0)^3 - ((21 - 7) * (MAX(Basic antral follicle count - 13, 0)^3) + (13 - 7) * (MAX(Basic antral follicle count - 21, 0)^3)) / (21 - 13)) / 324.

[0076] Basal antral follicle count 3 = (MAX(Basal antral follicle count - 10, 0)^3 - ((21 - 10) * (MAX(Basal antral follicle count - 13, 0)^3) + (13 - 10) * (MAX(Basal antral follicle count - 21, 0)^3)) / (21 - 13)) / 324.

[0077] Basal anti-Müllerian hormone 1 = (MAX(Basal anti-Müllerian hormone - 0.66, 0)^3 - ((8.76 - 0.66) * (MAX(Basal anti-Müllerian hormone - 4.3152, 0)^3) + (4.3152 - 0.66) * (MAX(Basal anti-Müllerian hormone - 8.76, 0)^3)) / (8.76 - 4.3152)) / 65.

[0078] Basal anti-Müllerian hormone 2 = (MAX(Basal anti-Müllerian hormone - 1.77, 0)^3 - ((8.76 - 1.77) * (MAX(Basal anti-Müllerian hormone - 4.3152, 0)^3) + (4.3152 - 1.77) * (MAX(Basal anti-Müllerian hormone - 8.76, 0)^3)) / (8.76 - 4.3152)) / 65.

[0079] Basal anti-Müllerian hormone 3 = (MAX(Basal anti-Müllerian hormone - 2.82, 0)^3 - ((8.76 - 2.82) * (MAX(Basal anti-Müllerian hormone - 4.3152, 0)^3) + (4.3152 - 2.82) * (MAX(Basal anti-Müllerian hormone - 8.76, 0)^3)) / (8.76 - 4.3152)) / 65.

[0080] Basal estrogen 1 = (MAX(Basal estrogen - 17, 0)^3 - ((84.19247 - 17) * (MAX(Basal estrogen - 53, 0)^3) + (53 - 17) * (MAX(Basal estrogen - 84.19247, 0)^3)) / (84.19247 - 53)) / 4514.

[0081] Basal estrogen 2 = (MAX(Basal estrogen - 30.98, 0)^3 - ((84.19247 - 30.98) * (MAX(Basal estrogen - 53, 0)^3) + (53 - 30.98) * (MAX(Basal estrogen - 84.19247, 0)^3)) / (84.19247 - 53)) / 4514.

[0082] Basal estrogen 3 = (MAX(basal estrogen - 40, 0)^3 - ((84.19247 - 40)*(MAX(basal estrogen - 53, 0)^3) + (53 - 40)*(MAX(basal estrogen - 84.19247, 0)^3)) / (84.19247 - 53)) / 4514.

[0083] Basal follicle stimulating hormone 1 = (MAX(basal follicle stimulating hormone - 4.64, 0)^3 - ((11.67 - 4.64)*(MAX(basal follicle stimulating hormone - 8.22, 0)^3) + (8.22 - 4.64)*(MAX(basal follicle stimulating hormone - 11.67, 0)^3)) / (11.67 - 8.22)) / 49.42.

[0084] Basal follicle stimulating hormone 2 = (MAX(basal follicle stimulating hormone - 6.07, 0)^3 - ((11.67 - 6.07)*(MAX(basal follicle stimulating hormone - 8.22, 0)^3) + (8.22 - 6.07)*(MAX(basal follicle stimulating hormone - 11.67, 0)^3)) / (11.67 - 8.22)) / 49.42.

[0085] Basal follicle stimulating hormone 3 = (MAX(basal follicle stimulating hormone - 7.03, 0)^3 - ((11.67 - 7.03)*(MAX(basal follicle stimulating hormone - 8.22, 0)^3) + (8.22 - 7.03)*(MAX(basal follicle stimulating hormone - 11.67, 0)^3)) / (11.67 - 8.22)) / 49.42.

[0086] Basal luteinizing hormone 1 = (MAX(basal luteinizing hormone - 2.05, 0)^3 - ((9.89 - 2.05)*(MAX(basal luteinizing hormone - 5.62, 0)^3) + (5.62 - 2.05)*(MAX(basal luteinizing hormone - 9.89, 0)^3)) / (9.89 - 5.62)) / 61.47.

[0087] Basal luteinizing hormone 2 = (MAX(basal luteinizing hormone - 3.35, 0)^3 - ((9.89 - 3.35)*(MAX(basal luteinizing hormone - 5.62, 0)^3) + (5.62 - 3.35)*(MAX(basal luteinizing hormone - 9.89, 0)^3)) / (9.89 - 5.62)) / 61.47.

[0088] Basal luteinizing hormone 3 = (MAX(Basal luteinizing hormone - 4.34, 0)^3 - ((9.89 - 4.34)*(MAX(Basal luteinizing hormone - 5.62, 0)^3) + (5.62 - 4.34)*(MAX(Basal luteinizing hormone - 9.89, 0)^3)) / (9.89 - 5.62)) / 61.47.

[0089] Basal progesterone 1 = (MAX(Basal progesterone - 0.19, 0)^3 - ((1.79 - 0.19)*(MAX(Basal progesterone - 0.9, 0)^3) + (0.9 - 0.19)*(MAX(Basal progesterone - 1.79, 0)^3)) / (1.79 - 0.9)) / 2.56.

[0090] Basal progesterone 2 = (MAX(Basal progesterone - 0.43, 0)^3 - ((1.79 - 0.43)*(MAX(Basal progesterone - 0.9, 0)^3) + (0.9 - 0.43)*(MAX(Basal progesterone - 1.79, 0)^3)) / (1.79 - 0.9)) / 2.56.

[0091] Basal progesterone 3 = (MAX(Basal progesterone - 0.62, 0)^3 - ((1.79 - 0.62)*(MAX(Basal progesterone - 0.9, 0)^3) + (0.9 - 0.62)*(MAX(Basal progesterone - 1.79, 0)^3)) / (1.79 - 0.9)) / 2.56.

[0092] Basal prolactin 1 = (MAX(Basal prolactin - 6.54, 0)^3 - ((31.538 - 6.54)*(MAX(Basal prolactin - 18.8, 0)^3) + (18.8 - 6.54)*(MAX(Basal prolactin - 31.538, 0)^3)) / (31.538 - 18.8)) / 624.9.

[0093] Basal prolactin 2 = (MAX(Basal prolactin - 10.79, 0)^3 - ((31.538 - 10.79)*(MAX(Basal prolactin - 18.8, 0)^3) + (18.8 - 10.79)*(MAX(Basal prolactin - 31.538, 0)^3)) / (31.538 - 18.8)) / 624.9.

[0094] Basal prolactin 3 = (MAX(Basal prolactin - 14.18, 0)^3 - ((31.538 - 14.18)*(MAX(Basal prolactin - 18.8, 0)^3) + (18.8 - 14.18)*(MAX(Basal prolactin - 31.538, 0)^3)) / (31.538 - 18.8)) / 624.9.

[0095] Cleavage rate 1 = (MAX(Cleavage rate - 38.17308, 0)^3 - ((97.4359 - 38.17308)*(MAX(Cleavage rate - 93.75, 0)^3)+(93.75 - 38.17308)*(MAX(Cleavage rate - 97.4359, 0)^3)) / (97.4359 - 93.75)) / 3504.41.

[0096] Cleavage rate 2 = (MAX(Cleavage rate - 83.33333, 0)^3 - ((97.4359 - 83.33333)*(MAX(Cleavage rate - 93.75, 0)^3)+(93.75 - 83.33333)*(MAX(Cleavage rate - 97.4359, 0)^3)) / (97.4359 - 93.75)) / 3504.41.

[0097] Cleavage rate 3 = (MAX(Cleavage rate - 88.88889, 0)^3 - ((97.4359 - 88.88889)*(MAX(Cleavage rate - 93.75, 0)^3)+(93.75 - 88.88889)*(MAX(Cleavage rate - 97.4359, 0)^3)) / (97.4359 - 93.75)) / 3504.41.

[0098] Number of cleavage 1 = (MAX(Number of cleavage - 1, 0)^3 - ((15 - 1)*(MAX(Number of cleavage - 9, 0)^3)+(9 - 1)*(MAX(Number of cleavage - 15, 0)^3)) / (15 - 9)) / 196.

[0099] Number of cleavage 2 = (MAX(Number of cleavage - 4, 0)^3 - ((15 - 4)*(MAX(Number of cleavage - 9, 0)^3)+(9 - 4)*(MAX(Number of cleavage - 15, 0)^3)) / (15 - 9)) / 196.

[0100] Number of cleavage 3 = (MAX(Number of cleavage - 6, 0)^3 - ((15 - 6)*(MAX(Number of cleavage - 9, 0)^3)+(9 - 6)*(MAX(Number of cleavage - 15, 0)^3)) / (15 - 9)) / 196.

[0101] Infertility duration 1 = (MAX(Infertility duration - 0.8, 0)^3 - ((10 - 0.8)*(MAX(Infertility duration - 5, 0)^3)+(5 - 0.8)*(MAX(Infertility duration - 10, 0)^3)) / (10 - 5)) / 84.64.

[0102] Infertility duration 2 = (MAX(Infertility duration - 2, 0)^3 - ((10 - 2)*(MAX(Infertility duration - 5, 0)^3) + (5 - 2)*(MAX(Infertility duration - 10, 0)^3)) / (10 - 5)) / 84.64.

[0103] Infertility duration 3 = (MAX(Infertility duration - 3.5, 0)^3 - ((10 - 3.5)*(MAX(Infertility duration - 5, 0)^3) + (5 - 3.5)*(MAX(Infertility duration - 10, 0)^3)) / (10 - 5)) / 84.64.

[0104] Embryo number 1 = (MAX(Embryo number - 1, 0)^3 - ((13 - 1)*(MAX(Embryo number - 8, 0)^3) + (8 - 1)*(MAX(Embryo number - 13, 0)^3)) / (13 - 8)) / 144.

[0105] Embryo number 2 = (MAX(Embryo number - 3, 0)^3 - ((13 - 3)*(MAX(Embryo number - 8, 0)^3) + (8 - 3)*(MAX(Embryo number - 13, 0)^3)) / (13 - 8)) / 144.

[0106] Embryo number 3 = (MAX(Embryo number - 5, 0)^3 - ((13 - 5)*(MAX(Embryo number - 8, 0)^3) + (8 - 5)*(MAX(Embryo number - 13, 0)^3)) / (13 - 8)) / 144.

[0107] 2PN fertilization rate 1 = (MAX(2PN fertilization rate - 31.57895, 0)^3 - ((100 - 31.57895)*(MAX(2PN fertilization rate - 78.94737, 0)^3) + (78.94737 - 31.57895)*(MAX(2PN fertilization rate - 100, 0)^3)) / (100 - 78.94737)) / 4681.44.

[0108] 2PN fertilization rate 2 = (MAX(2PN fertilization rate - 54.54545, 0)^3 - ((100 - 54.54545)*(MAX(2PN fertilization rate - 78.94737, 0)^3) + (78.94737 - 54.54545)*(MAX(2PN fertilization rate - 100, 0)^3)) / (100 - 78.94737)) / 4681.44.

[0109] 2PN fertilization rate 3 = (MAX(2PN fertilization rate - 66.66667, 0)^3 - ((100 - 66.66667)*(MAX(2PN fertilization rate - 78.94737, 0)^3) + (78.94737 - 66.66667)*(MAX(2PN fertilization rate - 100, 0)^3)) / (100 - 78.94737)) / 4681.44。

[0110] Fertilization rate 1 = (MAX(fertilization rate - 46.15385, 0)^3 - ((100 - 46.15385)*(MAX(fertilization rate - 84.21053, 0)^3) + (84.21053 - 46.15385)*(MAX(fertilization rate - 100, 0)^3)) / (100 - 84.21053)) / 2899.41。

[0111] Fertilization rate 2 = (MAX(fertilization rate - 71.42857, 0)^3 - ((100 - 71.42857)*(MAX(fertilization rate - 84.21053, 0)^3) + (84.21053 - 71.42857)*(MAX(fertilization rate - 100, 0)^3)) / (100 - 84.21053)) / 2899.41。

[0112] Female age 1 = (MAX(female age - 25, 0)^3 - ((39 - 25)*(MAX(female age - 34, 0)^3) + (34 - 25)*(MAX(female age - 39, 0)^3)) / (39 - 34)) / 196。

[0113] Female age 2 = (MAX(female age - 29, 0)^3 - ((39 - 29)*(MAX(female age - 34, 0)^3) + (34 - 29)*(MAX(female age - 39, 0)^3)) / (39 - 34)) / 196。

[0114] Female age 3 = (MAX(female age - 31, 0)^3 - ((39 - 31)*(MAX(female age - 34, 0)^3) + (34 - 31)*(MAX(female age - 39, 0)^3)) / (39 - 34)) / 196。

[0115] Female body mass index 1 = (MAX(female body mass index - 17.6, 0)^3 - ((24.6 - 17.6)*(MAX(female body mass index - 22.6, 0)^3) + (22.6 - 17.6)*(MAX(female body mass index - 24.6, 0)^3)) / (24.6 - 22.6)) / 49。

[0116] Female Body Mass Index 2 = (MAX(Female Body Mass Index - 19.7, 0)^3 - ((24.6 - 19.7) * (MAX(Female Body Mass Index - 22.6, 0)^3) + (22.6 - 19.7) * (MAX(Female Body Mass Index - 24.6, 0)^3)) / (24.6 - 22.6)) / 49.

[0117] Female Body Mass Index 3 = (MAX(Female Body Mass Index - 21.11, 0)^3 - ((24.6 - 21.11) * (MAX(Female Body Mass Index - 22.6, 0)^3) + (22.6 - 21.11) * (MAX(Female Body Mass Index - 24.6, 0)^3)) / (24.6 - 22.6)) / 49.

[0118] Female Height 1 = (MAX(Female Height - 150, 0)^3 - ((167 - 150) * (MAX(Female Height - 161, 0)^3) + (161 - 150) * (MAX(Female Height - 167, 0)^3)) / (167 - 161)) / 289.

[0119] Female Height 2 = (MAX(Female Height - 155.5, 0)^3 - ((167 - 155.5) * (MAX(Female Height - 161, 0)^3) + (161 - 155.5) * (MAX(Female Height - 167, 0)^3)) / (167 - 161)) / 289.

[0120] Female Height 3 = (MAX(Female Height - 158, 0)^3 - ((167 - 158) * (MAX(Female Height - 161, 0)^3) + (161 - 158) * (MAX(Female Height - 167, 0)^3)) / (167 - 161)) / 289.

[0121] Female Weight 1 = (MAX(Female Weight - 43, 0)^3 - ((64 - 43) * (MAX(Female Weight - 57, 0)^3) + (57 - 43) * (MAX(Female Weight - 64, 0)^3)) / (64 - 57)) / 441.

[0122] Female Weight 2 = (MAX(Female Weight - 49, 0)^3 - ((64 - 49) * (MAX(Female Weight - 57, 0)^3) + (57 - 49) * (MAX(Female Weight - 64, 0)^3)) / (64 - 57)) / 441.

[0123] Female weight 3 = (MAX(Female weight - 53, 0)^3 - ((64 - 53)*(MAX(Female weight - 57, 0)^3) + (57 - 53)*(MAX(Female weight - 64, 0)^3)) / (64 - 57)) / 441.

[0124] Fertilized egg number 1 = (MAX(Fertilized egg number - 2, 0)^3 - ((17 - 2)*(MAX(Fertilized egg number - 10, 0)^3) + (10 - 2)*(MAX(Fertilized egg number - 17, 0)^3)) / (17 - 10)) / 225.

[0125] Fertilized egg number 2 = (MAX(Fertilized egg number - 5, 0)^3 - ((17 - 5)*(MAX(Fertilized egg number - 10, 0)^3) + (10 - 5)*(MAX(Fertilized egg number - 17, 0)^3)) / (17 - 10)) / 225.

[0126] Fertilized egg number 3 = (MAX(Fertilized egg number - 7, 0)^3 - ((17 - 7)*(MAX(Fertilized egg number - 10, 0)^3) + (10 - 7)*(MAX(Fertilized egg number - 17, 0)^3)) / (17 - 10)) / 225.

[0127] High-quality embryo number 1 = (MAX(High-quality embryo number - 0, 0)^3 - ((10 - 0)*(MAX(High-quality embryo number - 5, 0)^3) + (5 - 0)*(MAX(High-quality embryo number - 10, 0)^3)) / (10 - 5)) / 100.

[0128] High-quality embryo number 2 = (MAX(High-quality embryo number - 2, 0)^3 - ((10 - 2)*(MAX(High-quality embryo number - 5, 0)^3) + (5 - 2)*(MAX(High-quality embryo number - 10, 0)^3)) / (10 - 5)) / 100.

[0129] High-quality embryo number 3 = (MAX(High-quality embryo number - 3, 0)^3 - ((10 - 3)*(MAX(High-quality embryo number - 5, 0)^3) + (5 - 3)*(MAX(High-quality embryo number - 10, 0)^3)) / (10 - 5)) / 100.

[0130] Male age 1 = (MAX(Male age - 26, 0)^3 - ((42 - 26)*(MAX(Male age - 36, 0)^3) + (36 - 26)*(MAX(Male age - 42, 0)^3)) / (42 - 36)) / 256.

[0131] Male age 2 = (MAX(Male age - 30, 0)^3 - ((42 - 30)*(MAX(Male age - 36, 0)^3) + (36 - 30)*(MAX(Male age - 42, 0)^3)) / (42 - 36)) / 256.

[0132] Male age 3 = (MAX(Male age - 32, 0)^3 - ((42 - 32)*(MAX(Male age - 36, 0)^3) + (36 - 32)*(MAX(Male age - 42, 0)^3)) / (42 - 36)) / 256.

[0133] Male BMI 1 = (MAX(Male BMI - 18.73, 0)^3 - ((29.74 - 18.73)*(MAX(Male BMI - 25.71, 0)^3) + (25.71 - 18.73)*(MAX(Male BMI - 29.74, 0)^3)) / (29.74 - 25.71)) / 121.

[0134] Male BMI 2 = (MAX(Male BMI - 21.72, 0)^3 - ((29.74 - 21.72)*(MAX(Male BMI - 25.71, 0)^3) + (25.71 - 21.72)*(MAX(Male BMI - 29.74, 0)^3)) / (29.74 - 25.71)) / 121.

[0135] Male BMI 3 = (MAX(Male BMI - 23.67, 0)^3 - ((29.74 - 23.67)*(MAX(Male BMI - 25.71, 0)^3) + (25.71 - 23.67)*(MAX(Male BMI - 29.74, 0)^3)) / (29.74 - 25.71)) / 121.

[0136] Male height 1 = (MAX(Male height - 162, 0)^3 - ((180 - 162)*(MAX(Male height - 174, 0)^3) + (174 - 162)*(MAX(Male height - 180, 0)^3)) / (180 - 174)) / 324.

[0137] Male height 2 = (MAX(Male height - 168, 0)^3 - ((180 - 168)*(MAX(Male height - 174, 0)^3) + (174 - 168)*(MAX(Male height - 180, 0)^3)) / (180 - 174)) / 324.

[0138] Male height 3 = (MAX(Male height - 170, 0)^3 - ((180 - 170)*(MAX(Male height - 174, 0)^3) + (174 - 170)*(MAX(Male height - 180, 0)^3)) / (180 - 174)) / 324.

[0139] Mature egg rate 1 = (MAX(Mature egg rate - 57.14286, 0)^3 - ((97.56098 - 57.14286)*(MAX(Mature egg rate - 94.11765, 0)^3) + (94.11765 - 57.14286)*(MAX(Mature egg rate - 97.56098, 0)^3)) / (97.56098 - 94.11765)) / 1633.62.

[0140] Mature egg rate 2 = (MAX(Mature egg rate - 80, 0)^3 - ((97.56098 - 80)*(MAX(Mature egg rate - 94.11765, 0)^3) + (94.11765 - 80)*(MAX(Mature egg rate - 97.56098, 0)^3)) / (97.56098 - 94.11765)) / 1633.62.

[0141] Mature egg rate 3 = (MAX(Mature egg rate - 87.5, 0)^3 - ((97.56098 - 87.5)*(MAX(Mature egg rate - 94.11765, 0)^3) + (94.11765 - 87.5)*(MAX(Mature egg rate - 97.56098, 0)^3)) / (97.56098 - 94.11765)) / 1633.62.

[0142] Retrieved oocyte number 1 = (MAX(Retrieved oocyte number - 2, 0)^3 - ((21 - 2)*(MAX(Retrieved oocyte number - 13, 0)^3) + (13 - 2)*(MAX(Retrieved oocyte number - 21, 0)^3)) / (21 - 13)) / 361.

[0143] Retrieved oocyte number 2 = (MAX(Retrieved oocyte number - 6, 0)^3 - ((21 - 6)*(MAX(Retrieved oocyte number - 13, 0)^3) + (13 - 6)*(MAX(Retrieved oocyte number - 21, 0)^3)) / (21 - 13)) / 361.

[0144] Retrieved oocyte number 3 = (MAX(Retrieved oocyte number - 9, 0)^3 - ((21 - 9)*(MAX(Retrieved oocyte number - 13, 0)^3) + (13 - 9)*(MAX(Retrieved oocyte number - 21, 0)^3)) / (21 - 13)) / 361.

[0145] In practical applications, in step S1, the criteria for confirming the inclusion of subjects are to exclude any contraindications to assisted reproductive technology and to be patients eligible for ET.

[0146] In practical applications, in step S2, variables are selected based on the Least Absolute Shrinkage and Selection Operator (Lasso) method. Lasso selects variables by introducing a penalty term into the multiple linear regression model and can be written as:

[0147] λΣ|βj|

[0148] Where: Σ: The Greek symbol represents the sum; j ranges from 1 to the number of variables; β is the regression coefficient of the jth variable; λ is a constant greater than or equal to 0 used to control the intensity of the penalty term. When λ = 0, all variables are included, and when λ increases, the intensity of the penalty term increases and more variables are excluded.

[0149] The selection of λ is achieved by generating the lowest possible test Mean Squared Error (MSE). MSE is defined as MSE = (1 / n)*Σ(yi - f(xi))2.

[0150] Where: n is the total number of observations; i is a natural number from 1 to n; yi is the response value of the ith observation; f(xi) is the predicted response value of the ith observation.

[0151] Preferably, the R software package glmnet is used. By randomly splitting the training set into a training set and a validation set at a ratio of 9:1 and performing 10-fold cross-validation, the test MSE is the average of the MSEs in the 10-fold cross-validation, and the λ value when the test MSE is minimized is obtained.

[0152] After substituting the λ value into the model, the regression coefficient of each variable can be obtained. Variables with a regression coefficient of 0 are excluded from the model. The final obtained regression coefficients and intercept can be used to calculate the predicted value. After determining λ through cross-validation, for a given λ solution, the LASSO model can, like a generalized linear model, give an equation containing the regression coefficients and intercepts of each variable for the next prediction. The values of the intercept and coefficients vary according to the solution of the specific data and are specifically described in the modeling. The constructed models each have unique regression coefficients and intercepts. The intercept is the constant term in the model that is independent of the variable size. Taking the following non-gestational sac model as an example:

[0153] y = 1.45577 + fresh non-downregulated cycle * (0.2264998) + frozen-thawed downregulated cycle

[0154] * (0.5737309) + …….

[0155] 1.45577 is the intercept. 0.2264998 is the regression coefficient for the variable / feature "fresh non - down - regulated cycle", and so on.

[0156] In practical applications, in step S3, a model is built based on the variables finally selected in step S2. y = β1X1+…+βiXi + intercept. Where i is the i - th variable selected in step S2; β is the coefficient of the i - th variable.

[0157] The details of the model are as follows:

[0158] No gestational sac

[0159] y = 1.45577 + Fresh non - down - regulated cycle * (0.2264998) + Frozen - thawed down - regulated cycle * (0.5737309) + Frozen - thawed hormone - replacement cycle * (0.427264) + Frozen - thawed natural cycle * (0.6730623) + Frozen - thawed induced - ovulation cycle * (-0.05673244) + Poor endometrial echo morphology * (0.1796946) + Distance from transplantation to uterine fundus * (0.2718966) + Single blastomere transplantation * (0.5412932) + [Number of transplantations = 2] * (0.06639986) + [Number of transplantations > 2] * (0.1762219) + Quality of transplanted embryo (ordinary) * (0.2453566) + Quality of transplanted embryo (poor) * (1.061594) + Quality of transplanted embryo (good + ordinary) * (0.1814903) + Quality of transplanted embryo (good + poor) * (0.03814809) + Quality of transplanted embryo (ordinary + poor) * (0.3936408) + Intracytoplasmic sperm injection * (-0.004600189) + Whole - embryo cryopreservation * (-0.08506118) + Duration of infertility * (0.03404056) + Polycystic ovary syndrome * (-0.08676948) + Abnormal uterine morphology * (0.1652612) + Intrauterine adhesions * (0.1227428) + Previous live birth * (-0.1929755) + Female weight * (-0.01076688) + Female body mass index * (-0.004670255) + Female follicle - stimulating hormone * (0.003929719) + Basal luteinizing hormone * (-0.0000528649) + Basal prolactin * (0.00002650024) + Basal progesterone * (0.009816736) + Basal anti - Müllerian hormone * (-0.0055641) + Basal antral follicle count * (-0.00857774) + Male height * (-0.003206887) + Male body mass index * (0.00602765) + 2PN fertilization rate * (-0.001618058) + Number of MI oocytes * (0.02207378) + Number of degenerated oocytes * (0.002447852) + Number of embryos * (-0.01407337) + Number of good - quality embryos * (-0.01010365) + Cesarean section history * (0.07424005) + Uterine scar diverticulum * (0.03081223) + Starting dose * (0.001862502) + Female age 3 * (0.3529343) + Female body mass index 3 * (0.1747765) + Basal follicle - stimulating hormone 2 * (0.003430329) + Basal follicle - stimulating hormone 3 * (0.008503291) + Basal luteinizing hormone 1 * (-0.004563992) + Basal estrogen 3 * (0.001830259) + Basal antral follicle count 3 * (0.Number of retrieved oocytes 2 * (-0.006352761) + Maturation rate of oocytes 1 * (0.002452197) + Fertilization rate 2 * (-0.009546101) + Number of cleaved embryos 1 * (0.00513727) + Cleavage rate 1 * (-0.003481237) + Number of high-quality embryos 2 * (0.02958512).

[0160] Single gestational sac

[0161] y = 1.178562 + Double blastocyst transfer * (-0.102401) + Double pronuclear transfer * (-0.3053187) + Intracytoplasmic sperm injection * (0.01825063) + Cryopreservation of all embryos * (0.02900404) + Endometriosis * (0.07162082) + Uterine myoma * (-0.008477006) + Previous live birth * (0.01278788) + Previous spontaneous abortion * (-0.03704392) + Basal estrogen * (-0.00005904638) + Basal progesterone * (-0.001602434) + Basal testosterone * (0.003282056) + Male weight * (-0.0001915054) + Male body mass index * (-0.000337807) + Number of mature oocytes * (-0.002070243) + Maturation rate of oocytes * (-0.001084511) + Number of germinal vesicle oocytes * (-0.01375337) + Number of degenerated oocytes * (-0.02254867) + History of cesarean section * (-0.02893249) + Female age * (0.001369586) + Female age 1 * (0.002129808) + Male age 1 * (0.001163889) + Duration of infertility 3 * (0.06888685) + Female weight 1 * (-0.001851646) + Basal follicle-stimulating hormone 3 * (-0.01021028) + Basal prolactin 3 * (-0.00118734) + Basal antral follicle count 3 * (-0.006752027) + Male height 3 * (-0.005945586) + Male weight 3 * (-0.005106048) + Male body mass index 2 * (0.01102081) + Male body mass index 3 * (0.0005279188) + Number of fertilized eggs 3 * (0.002146346) + Fertilization rate 2 * (0.000584984) + 2PN fertilization rate 3 * (0.0002873137) + Cleavage rate 3 * (-0.3697333) + Number of embryos 3 * (0.0003284333)

[0162] Multiple gestational sacs

[0163] y = -2.6343317528 + Fresh non - down - regulated cycle * (-0.3707699169) + Frozen - thawed down - regulated cycle * (-0.5791721826) + Frozen - thawed hormone - replacement cycle * (-0.4846289902) + Frozen - thawed natural cycle * (-0.4795406563) + Frozen - thawed induced - ovulation cycle * (0.0210331198) + Poor endometrial echo morphology * (-0.0296376773) + Double blastocyst transfer * (3.2801768547) + Single blastomere transfer * (-0.6644765381) + Double blastomere transfer * (2.5787172464) + [Number of transfers = 2] * (-0.0690059938) + [Number of transfers > 2] * (-0.152523982) + Transferred embryo quality (ordinary) * (-0.0367136845) + Transferred embryo quality (poor) * (-0.7635627856) + Transferred embryo quality (good + ordinary) * (-0.0118720391) + Transferred embryo quality (good + poor) * (-0.6228959489) + Transferred embryo quality (ordinary + poor) * (-0.5341987389) + Endometriosis * (-0.0618362862) + Uterine morphological abnormality * (-0.1822646724) + Uterine myoma * (0.1597235576) + Female weight * (0.0029374666) + Basal testosterone * (-0.0387989825) + Basal antral follicle count * (0.0070576979) + 2PN fertilization rate * (0.0021605145) + MI oocyte count * (-0.0099029967) + GV oocyte count * (0.0163909409) + Good embryo count * (0.0633175711) + Uterine scar diverticulum * (-0.133120707) + Starting dose * (-0.0003340564) + Male age * (-0.0010576355) + Female age 1 * (-0.0337000291) + Female age 2 * (-0.0296723787) + Female age 3 * (-0.0681963877) + Female weight 1 * (0.006667733) + Female weight 3 * (-0.0038118422) + Female body mass index 3 * (-0.0117098964) + Basal luteinizing hormone 1 * (0.0008389895) + Basal anti - Müllerian hormone 3 * (-0.0174795304) + 2PN fertilization rate 3 * (-0.003150815) + Good embryo count 2 * (-0.0599028865) + Good embryo count 3 * (-0.0825791609)

[0164] According to the link function predicted by the model, the y value is converted into a predicted value. The final predicted value is p = exp(y) / (1 - exp(y)).

[0165] This method is tested and calibrated in the validation set. The Receiver Operator Characteristic (ROC) curve is used to test the discriminatory power of the model in the validation set. Calibration curves are used to test the consistency between the predicted values and the observed values of the prediction model in the validation set. The specific method is to use the predicted value as the abscissa and the observed value as the ordinate, and perform linear and smooth fitting on each individual data (observation) in the validation set. According to the results of the linear fitting, the slope and intercept of the fitting curve are obtained. The predicted value is calibrated according to the slope and intercept. The method is pˋ = slope * p + intercept. Where pˋ is the calibrated predicted value, p is the predicted value before calibration; slope is the slope obtained from the linear fitting of the above calibration curve; intercept is the intercept obtained from the linear fitting of the above calibration curve. The calibrated predicted value is re-smoothed and fitted with the observed value. The criterion for judging whether the fitting is accurate is the consistency with the straight line with a slope of 1 and an intercept of 0 in the coordinate system.

[0166] In practical applications, in step S5, the online tool uses a traditional web construction framework. The front end uses JQuery, Bootstrap, JavaScript, and HTML to write the basic interaction logic and user interface, uses echarts to plot and visualize, and the back end uses the Djiango framework of Python3 to write the preprocessing of network requests and the prediction of the model. After writing, as Figure 1 shown, use the data input module to obtain patient medical record data or manually input the patient information specified in S1; input the parameters specified in S1 according to clinical needs. Perform the specified conversion (dummy variable, cubic spline) on the input information according to the method of S1. The data prediction module generates a predicted outcome according to the model formula established in S2-3 and the results of the input module. The user interface module is used to display the predicted results, adjust the prediction data, and provide feedback. The embryo transfer outcome prediction method provided by this application can be applied in predicting ET outcomes and multiple pregnancy risks, and can predict whether the outcome of embryo transfer is failure, singleton pregnancy, or multiple pregnancy based on variables such as the patient's clinical parameters, treatment conditions, morphology, stage, and number of transplanted embryos. The outcome is defined by the number of gestational sacs under transvaginal ultrasound.

[0167] Verify the prediction performance of the model

[0168] Two methods, namely LASSO (this model) and XGboost, were used for modeling respectively. The modeling training data came from the transplantation cycles of the Reproductive Medicine Center from 2013 to 2020; the validation set came from the transplantation cycles from 2021 to 2022, which was a time-series external validation. A total of 45,473 transplantation cycles were included, with 39,980 cycles used for modeling training; 5,493 cycles were used for time-series validation. The predictive ability of the model was evaluated using the Receiver Operating Characteristic Curve (ROC) and its Area Under the Curve (ROCAUC). The overall predictive performance of the model was evaluated using micro ROC and macro ROC respectively. In the training data, both LASSO (micro ROCAUC 0.76, 95% CI: 0.75, 0.76; macro ROCAUC 0.73, 95% CI: 0.72, 0.73) and XGboost (micro ROCAUC 0.8, 95% CI: 0.80, 0.81; macro ROCAUC 0.78, 95% CI: 0.781, 0.784) showed moderate predictive value, and the predictive ability of the XGboost method was slightly higher than that of the LASSO method. However, in the validation set, the predictive abilities of the two were similar (micro ROCAUC 0.78, 95% CI: 0.77, 0.79; macro ROCAUC 0.73, 95% CI: 0.72, 0.74). The macroROC and microROC curves are shown in Figures 2 - 5 . In the figure, the XGboost method is represented by xgb, and LASSO is represented by las.

[0169] The classification performance of each outcome of the three-category variable (no gestational sac implantation, single gestational sac implantation, multiple gestational sac implantations) is shown in Table 1, which shows the sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of each model in the validation set.

[0170] Table 1 Predictive performance of the model for no gestational sac implantation, single gestational sac implantation, and multiple gestational sac implantations

[0171]

[0172] Comparison with existing models

[0173] Multiple gestational sac implantation is a major risk to be considered during the ET process. Existing models for predicting multiple gestational sac implantation (multiple pregnancy) mainly come from Luke et al. (http: / / dx.doi.org / 10.1016 / j.fertnstert.2014.05.020) and are trained on the American Society for Reproductive Medicine database. In addition, clinically, the combination of the female patient and the number of transferred embryos is often used to roughly estimate the risk of multiple gestational sac implantation. Based on the above data (training data + validation data), the discriminatory power of the same model and traditional evaluation indicators (age, number of transferred embryos) for multiple gestational sac implantation was compared. Figure 6 The comparison of the ROC curves and the areas under the ROC curves is shown. The results indicate that the discriminatory power of the model of this application is significantly higher than that of the Luke model and the combination of the female patient and the number of transferred embryos, mainly because more clinical variables are incorporated.

[0174] Contribution to clinical decision-making

[0175] The potential impact of different models and traditional evaluation indicators (age, number of transferred embryos) and existing models (http: / / dx.doi.org / 10.1016 / j.fertnstert.2014.05.020) on decision-making was analyzed through the Decision Curve Analysis (DCA). The abscissa of the DCA represents the threshold probability, which represents the decision-making preference or psychological threshold of different patients or clinicians, indicating the probability at which the decision-maker believes the benefits and risks of the treatment behavior are balanced. In this case, it represents the acceptable threshold for multiple gestational sac implantation. As Figure 7 shown, the ordinate is the net benefit, which represents the true positive prediction minus the false positive weighted by the threshold probability. The net benefit curves of different models represent the contribution of model predictions to decision-making at different threshold probabilities. As long as the net benefit is positive at a given threshold, it can be considered that the model prediction contributes to the decision-making. The leftmost curve represents maintaining the original embryo transfer strategy without prediction, and it can be seen that this strategy is only useful for patients with a threshold < 13.7%. Using the prediction model can assist more patients with different decision-making preferences or psychological thresholds in making decisions. Compared with the Luke model or traditional evaluation indicators (age, number of transferred embryos), this model can provide a basis for decision-making for a wider group of people because it can cover a wider range of threshold probabilities.

[0176] The present application also provides an application scenario, which applies the above-mentioned embryo transfer outcome prediction method. Specifically: The embryo transfer outcome prediction method provided in this embodiment can be applied in the consultation scenario before clinical embryo transfer. The consultation scenario includes the acquisition of clinical information, the processing of clinical information, and the distribution link of decision-making consultation information; the acquisition of clinical information refers to the information obtained through relevant diagnoses and treatments before making a decision (here the decision refers to embryo transfer). The obtained clinical information obtains consultation information (including: single-gestational sac implantation, multiple-gestational sac implantation, and no-gestational sac implantation) to support the decision through the information processing link. The distribution of decision-making consultation information can be distributed through a human-machine interface or by a clinician explaining to the patient.

[0177] Based on the same inventive concept, the embodiment of the present application also provides a video tag processing device of an embryo transfer outcome prediction device for implementing the above-mentioned embryo transfer outcome prediction method. The solution provided by this device to solve the problem is similar to the solution described in the above method. Therefore, the specific limitations in one or more of the following embodiments of the embryo transfer outcome prediction device can refer to the limitations on the embryo transfer outcome prediction method in the above text, and will not be repeated here.

[0178] In an exemplary embodiment, an embryo transfer outcome prediction device is provided, including:

[0179] An acquisition module, configured to acquire historical transfer data of an embryo transfer patient; the historical transfer data includes historical clinical characteristic variables and historical embryo transfer outcomes; the historical clinical characteristic variables include continuous variables and categorical variables; the historical embryo transfer outcomes include transfer failure, singleton pregnancy, or multiple pregnancy.

[0180] A screening and partitioning module, configured to screen and partition the historical transfer data to obtain a data set; the data set includes a training set and an external validation set.

[0181] A feature screening module, configured to screen features of the data set to obtain risk factors.

[0182] A modeling module, configured to perform modeling using the risk factors of the training set and the historical embryo transfer outcomes of the training set to obtain a prediction model.

[0183] An adjustment module, configured to adjust the model parameters of the prediction model using the external validation set to obtain an adjusted prediction model.

[0184] A prediction module, configured to perform prediction using the adjusted prediction model according to the clinical characteristic variables of a pre-embryo transfer patient to obtain an embryo transfer outcome.

[0185] 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 10 FIG. 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 embryo transfer outcome prediction data. 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 embryo transfer outcome prediction method.

[0186] Those skilled in the art can understand that Figure 10 the structure shown in FIG. 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 above method embodiments are implemented.

[0187] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the above method embodiments are implemented.

[0188] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the above method embodiments are implemented.

[0189] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with relevant regulations.

[0190] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above 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 embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memories. Non-volatile memories can include read-only memory (ROM), magnetic tapes, floppy disks, flash memories, optical memories, high-density embedded non-volatile memories, resistive random access memories (ReRAM), magnetoresistive random access memories (MRAM), ferroelectric random access memories (FRAM), phase change memories (PCM), graphene memories, etc. Volatile memories can include random access memory (RAM) or external cache memories, 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.

[0191] The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processors, graphics processors, digital signal processors, programmable logics, data processing logics based on quantum computing, etc., without limitation.

[0192] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise 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 to be within the scope described in this specification.

[0193] Specific examples are used in this article to elaborate on the principles and implementation manners of the present application. The descriptions of the above embodiments are 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 embryo transplantation outcome, characterized in that: The method for predicting the outcome of embryo transplantation comprises: Obtaining historical transplantation data of embryo transplant patients; the historical transplantation data includes historical clinical characteristic variables and historical embryo transplantation outcomes; the historical clinical characteristic variables include continuous variables and categorical variables; the historical embryo transplantation outcomes include transplantation failure, single pregnancy or multiple pregnancy; Screening and dividing the historical transplantation data to obtain a data set; the data set includes a training set and an external validation set; Performing feature screening on the data set to obtain risk factors; Modeling using the risk factors of the training set and the historical embryo transfer outcomes of the training set to obtain a prediction model; Using the external validation set to adjust the model parameters of the prediction model to obtain an adjusted prediction model; The adjusted prediction model is used to predict the embryo transfer outcome according to the clinical characteristic variables of the patients undergoing embryo transfer.

2. The method for predicting embryo transplantation outcome according to claim 1, characterized in that: The historical transplantation data is screened and divided to obtain a data set, specifically including: The historical transplantation data are screened according to the screening criteria to obtain screened data; the screening criteria are the historical transplantation data of patients who exclude contraindications to assisted reproductive technology and meet the requirements for embryo transplantation; Perform dummy variable conversion on the categorical variables in the screened data to obtain converted categorical variables; The continuous variables in the screened data are fitted with the historical embryo transfer outcomes using a generalized additive model to obtain variables with nonlinear relationships, and the variables with nonlinear relationships are transformed by restricted cubic strip transformation to obtain transformed continuous variables; The converted categorical variables and the converted continuous variables are divided into data to obtain a data set.

3. The method for predicting embryo transplantation outcome according to claim 1, characterized in that: The data set is subjected to feature screening to obtain risk factors, including: The least absolute shrinkage and selection operator methods are used to screen the features of the data set and obtain the risk factors.

4. The method for predicting embryo transplantation outcome according to claim 1, characterized in that: The expression of the prediction model is: y = β1X1 + … + βiX i + intercept; p = exp(y) / (1-exp(y)); Among them, y is the response value, βi is the coefficient of the i-th variable, p is the predicted value, X1 is the first related variable, X i is the i-th related variable.

5. The method for predicting embryo transplantation outcome according to claim 1, characterized in that: The continuous variables include female age, female height, female weight, female body mass index, basal follicle-stimulating hormone, basal luteinizing hormone, basal prolactin, basal estrogen, basal progesterone, basal testosterone, basal anti-Mullerian hormone, basal antral follicle number, male age, male height, male weight, male body mass index, infertility years, starting dose, number of retrieved eggs, number of mature eggs, mature egg rate, number of fertilized eggs, fertilization rate, number of 2PN fertilized eggs, 2PN fertilization rate, number of cleavages, number of 2PN cleavages, cleavage rate, number of MI eggs, number of GV eggs, number of degenerated eggs, number of embryos, number of high-quality embryos, endometrial thickness and distance from transplantation to the fundus.

6. The method for predicting embryo transplantation outcome according to claim 1, characterized in that: The categorical variables included previous live birth, previous spontaneous abortion, endometriosis, polycystic ovary syndrome, abnormal uterine morphology, uterine fibroids, intrauterine adhesions, history of cesarean section, endometrial scar diverticulum, intracytoplasmic sperm injection, cryopreservation, transfer cycle type, fresh down-regulation cycle, fresh non-down-regulation cycle, frozen-thawed down-regulation cycle, frozen-thawed hormone replacement cycle, frozen-thawed natural cycle, frozen-thawed induced ovulation cycle, poor endometrial echo morphology, transfer strategy, number of transfers, and transferred embryo morphology.

7. An embryo transplantation outcome prediction device, characterized in that: The embryo transplantation outcome prediction device comprises: An acquisition module is used to acquire historical transplantation data of embryo transplant patients; the historical transplantation data includes historical clinical characteristic variables and historical embryo transplantation outcomes; the historical clinical characteristic variables include continuous variables and categorical variables; the historical embryo transplantation outcomes include transplantation failure, single pregnancy or multiple pregnancy; A screening and division module, used for screening and dividing the historical transplantation data to obtain a data set; the data set includes a training set and an external verification set; A feature screening module, used to perform feature screening on the data set to obtain risk factors; A modeling module, used to perform modeling using the risk factors of the training set and the historical embryo transplantation outcomes of the training set to obtain a prediction model; An adjustment module, used to adjust the model parameters of the prediction model using the external validation set to obtain an adjusted prediction model; The prediction module is used to predict the outcome of embryo transplantation by using the adjusted prediction model according to the clinical characteristic variables of the patient undergoing embryo transplantation.

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 the outcome of embryo transplantation 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 the outcome of embryo transplantation 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 the outcome of embryo transplantation according to any one of claims 1 to 6 is implemented.

Images