A prediction model for embryo quality outcome of women of childbearing age undergoing assisted reproductive technology and a construction method thereof

Abstract

The present application relates to the field of biomedical technology, and more particularly to a prediction model for embryo quality outcome of women of childbearing age undergoing assisted reproductive technology; the present application is based on clinical and laboratory data of women of childbearing age undergoing assisted reproductive technology, uses light GBM multi-factor analysis to screen the key factors of clinical core parameters affecting women of childbearing age undergoing assisted reproductive technology, and performs spear correlation analysis with each clinical core parameter to obtain core factors affecting embryo quality outcome of women of childbearing age undergoing assisted reproductive technology, and to construct a model capable of accurately predicting embryo quality outcome, which can accurately predict the positive influencing factors of embryo quality, effectively reduce the psychological burden of a part of ART pregnant patients, so as to make clinical intervention more targeted, more conducive to the rational allocation of clinical resources, and suitable for large-scale popularization and application.

Description

Technical Field

[0001] This invention relates to the field of biomedical technology, specifically to a predictive model for embryo quality outcomes in women of reproductive age undergoing assisted reproductive technology. Background Technology

[0002] Assisted reproductive technology (ART) is a commonly used technique that provides more options for conception. Currently, there is an increasing trend of older pregnancies and births, leading to a growing preference for ART over natural conception. However, the safety of ART remains controversial. As mothers age, ovarian aging and declining ovarian reserve occur, and many older mothers often seek to improve pregnancy success rates by transferring multiple embryos, resulting in a significant increase in multiple pregnancies. Embryo quality is particularly important in ART pregnancy outcomes. Increasing the number of embryos allows patients more options; the total number of 2PN embryos and the number of high-quality embryos reflect the quality of the embryos transferred during ART.

[0003] Embryo quality may be related to factors such as the patient's oocyte quality, ovulation induction effect, and patient's physical condition, but currently, there is a lack of effective embryo quality prediction models. Therefore, there is an urgent clinical need for a simple and effective model and method that can assist doctors in accurately predicting the embryo quality outcome of assisted reproductive technologies (ART) in women of childbearing age based on the specific characteristics of ovulation induction parameters. This would accurately predict positive factors affecting embryo quality, thereby reducing the psychological burden on some patients undergoing ART, making clinical interventions more targeted, allocating clinical resources more rationally, and ultimately improving the success rate of ART. Summary of the Invention

[0004] The purpose of this invention is to provide a predictive model for embryo quality outcomes in women of reproductive age undergoing assisted reproductive technology and a method for constructing the model.

[0005] To achieve the above-mentioned objective, this invention provides a method for constructing a predictive model for embryo quality outcomes in women of reproductive age undergoing assisted reproductive technology, comprising the following steps:

[0006] S1. Obtain clinically relevant data of the sample; perform statistical analysis on the clinically relevant data of the sample to obtain statistical data of the sample; the statistical analysis includes: continuous data are expressed as mean (standard deviation), and categorical variables are expressed as frequency (percentage);

[0007] S2. Using the Light GBM model, key factors affecting the core clinical parameters were screened from the embryonic morphological parameters and patient baseline indicators of the samples. Spear correlation analysis was then performed between the key factors and each core clinical parameter, and the key factors with significant correlation were identified as the core factors of each core clinical parameter.

[0008] S3. The core factors of each clinical core parameter are re-incorporated into the light GBM model to construct a prediction model for each clinical core parameter.

[0009] S4. Evaluate the performance of the prediction model for each clinical core parameter, and optimize the prediction model based on the evaluation results to obtain a prediction model for embryo quality outcomes of women of childbearing age undergoing assisted reproductive technology.

[0010] This invention discloses a method for constructing a predictive model for embryo quality outcomes in women of reproductive age undergoing assisted reproductive technology (ART). Based on clinical and laboratory data from women of reproductive age undergoing ART, the method utilizes Light GBM multivariate analysis to screen key factors influencing embryo quality outcomes. Spear correlation analysis is then performed with each clinical core parameter to identify the core factors affecting embryo quality outcomes. Based on this, a model capable of accurately predicting embryo quality outcomes is constructed. This model can accurately predict positive influencing factors on embryo quality, effectively reducing the psychological burden on some patients undergoing ART, thereby making clinical intervention more targeted, facilitating the rational allocation of clinical resources, and suitable for large-scale application.

[0011] In step S1, preferably, the number of samples is not less than 1000.

[0012] Preferably, the clinically relevant data includes: embryo morphological parameters, patient baseline indicators, and core clinical parameters (used to measure embryo quality).

[0013] More preferably, the embryo morphological parameters include: cell number, cell uniformity (symmetry), cell fragmentation rate, embryonic development speed at the corresponding date, and multinucleation; the patient's baseline indicators include: AFC, BMI, baseline PRL, baseline T, baseline E2, total Gn medication dosage, HCG daily P, baseline FSH, HCG daily LH, baseline AMH, and HCG daily E2 for women of childbearing age; the clinical core parameters include the number of excellent embryos, the number of usable embryos, and the total number of 2PN embryos.

[0014] Preferably, the exclusion criteria for the sample include: multiple pregnancies (≥ triplets), women under 18 years of age or over 50 years of age at conception, incomplete or missing data on key variables (such as method of conception, maternal age, and neonatal outcome), women who conceived through intrauterine insemination (IUI), women with premature ovarian failure or diminished ovarian reserve (age ≤35 years and anti-Müllerian hormone [AMH] ≤1.1 ng / mL or antral follicle count [AFC] <5), and women whose male partners have azoospermia.

[0015] Preferably, R software 3.6.1 and SAS 9.4 are used to perform statistical analysis on the clinically relevant data of the samples.

[0016] Preferably, the Student's test is used for continuous data; for categorical variables expressed as frequencies (percentages), the χ² test or Fisher's exact test is used to compare differences between groups; error data are eliminated through tests.

[0017] In step S2, preferably, the Light GBM model includes at least 1000 decision trees; the more decision trees there are, the higher the accuracy of the constructed Light GBM model and the better the accuracy of the predicted model.

[0018] Preferably, the screening method for the key factors includes: sorting the embryonic morphological parameters and patient baseline indicators in descending order of their characteristic importance (Gain score), and selecting relevant factors with a total characteristic importance of not less than 70% as key factors.

[0019] In step S4, preferably, the evaluation method includes: (1) drawing a scatter plot of the predicted value and the true value, and overlaying a 1:1 reference line and a regression fitting line; (2) calculating the root mean square error RMSE.

[0020] Preferably, the scatter plot includes: a scatter plot of available embryos, a scatter plot of total 2PNs, and a scatter plot of good embryos.

[0021] Preferably, the criteria for verifying the predictive model are: R² value of the scatter plot ≥ 0.9; RMSE of the scatter plot ≥ 1.

[0022] Preferably, the optimization method for the prediction model includes increasing the number of samples.

[0023] To achieve the above-mentioned objectives, the present invention further provides a predictive model for embryo quality outcomes in women of reproductive age undergoing assisted reproductive technology, the predictive model being constructed using the method described above.

[0024] To further achieve the above-mentioned objectives, the present invention also provides an application of a predictive model for embryo quality outcomes in women of reproductive age undergoing assisted reproductive technology, used to predict embryo quality outcomes in women of reproductive age undergoing assisted reproductive technology.

[0025] Preferably, the application method includes: using the predictive model to visualize and interpret the embryo quality outcome of women of reproductive age undergoing assisted reproductive technology; specifically, SHAP analysis quantifies the global contribution of each feature to the embryo quality outcome through Summary Plot, Waterfall Plot reveals the driving path of single-sample prediction, and combines clinical thresholds to verify the nonlinear effects of key factors.

[0026] A device for predicting embryo quality outcomes in women of reproductive age undergoing assisted reproductive technology, the device being capable of running the aforementioned prediction model; the device can be used to predict embryo quality outcomes in women of reproductive age undergoing assisted reproductive technology.

[0027] Beneficial effects of the invention

[0028] 1. The present invention relates to a method for constructing a predictive model for embryo quality outcomes in women of reproductive age undergoing assisted reproductive technology. Based on clinical and laboratory data of women of reproductive age undergoing assisted reproductive technology, the method uses Light GBM multivariate analysis to screen key factors affecting embryo quality outcomes in women of reproductive age undergoing assisted reproductive technology, and performs Spear correlation analysis with clinical core parameters to obtain the core factors affecting embryo quality outcomes in women of reproductive age undergoing assisted reproductive technology, and constructs a model that can accurately predict embryo quality outcomes based on these factors.

[0029] 2. The predictive model for embryo quality outcomes in women of childbearing age undergoing assisted reproductive technology in this invention can accurately predict positive influencing factors on embryo quality, effectively reducing the psychological burden on some patients who conceive through ART.

[0030] 3. The predictive model of this invention is used to predict the embryo quality outcome of women of childbearing age undergoing assisted reproductive technology, making clinical intervention more targeted, more conducive to the rational allocation of clinical resources, and suitable for large-scale promotion and application. Attached Figure Description

[0031] Figure 1 This is a light GBM model of all factors affecting the number of usable embryos in Example 1 of the present invention;

[0032] Figure 2 This is a light GBM model representing all factors affecting the total number of 2PN in Embodiment 1 of the present invention;

[0033] Figure 3 This is a light GBM model of all factors affecting the number of high-quality embryos in Example 1 of the present invention;

[0034] Figure 4 These are all the key factors affecting the core clinical parameters in Embodiment 1 of the present invention;

[0035] Figure 5 Spear correlation between the number of high-quality embryos and key factors in Example 1 of this invention;

[0036] Figure 6 The correlation between the total number of 2PNs and key factors in Embodiment 1 of the present invention;

[0037] Figure 7The correlation between the number of available embryos and key factors in Example 1 of this invention;

[0038] Figure 8 This refers to the available embryo number prediction model constructed using core factors in Embodiment 1 of the present invention.

[0039] Figure 9 This refers to the 2PN total number prediction model constructed using core factors in Embodiment 1 of the present invention;

[0040] Figure 10 This is the high-quality embryo number prediction model constructed using core factors in Embodiment 1 of the present invention. Detailed Implementation

[0041] The technical solutions in some embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments provided in this disclosure are within the scope of protection of this disclosure.

[0042] Example 1

[0043] A predictive model for embryo quality outcomes in women of reproductive age undergoing assisted reproductive technology, the method of which includes the following steps:

[0044] S1. Obtain clinically relevant data from a sample (6371 women of childbearing age who underwent fresh cycle ART transplantation between 2013 and 2024); perform statistical analysis on the clinically relevant data of the sample (using R software 3.6.1 and SAS 9.4 to perform statistical analysis on the clinically relevant data of the sample), and obtain statistical data of the sample; the statistical analysis includes: continuous data are expressed as mean (standard deviation) (using Student's test), and categorical variables are expressed as frequency (percentage) (using χ2 test or Fisher's exact test to compare differences between groups).

[0045] Clinically relevant data (used to measure embryo quality) include: embryo morphological parameters, patient baseline indicators, and core clinical parameters;

[0046] Embryo morphological parameters include: cell number, cell uniformity (symmetry), cell fragmentation rate, embryonic development rate at the corresponding date, and multinucleation.

[0047] Patient baseline indicators include: AFC, BMI, baseline PRL, baseline T, baseline E2, total Gn medication dosage, HCG daily P, baseline FSH, HCG daily LH, baseline AMH, and HCG daily E2 for women of childbearing age.

[0048] Key clinical parameters include the number of high-quality embryos, the number of usable embryos, and the total number of 2PN embryos.

[0049] Exclusion criteria for the sample included: multiple pregnancies (≥ triplets), maternal age under 18 years or over 50 years, incomplete or missing data on key variables (such as method of conception, maternal age, and neonatal outcome), and cases conceived through intrauterine insemination (IUI) because its biological and procedural characteristics differ from in vitro fertilization / intracytoplasmic sperm injection-embryo transfer (IVF / ICSI-ET); women with premature ovarian failure or diminished ovarian reserve (age ≤35 years and anti-Müllerian hormone [AMH] ≤1.1 ng / mL or antral follicle count [AFC] <5) or male partners with azoospermia, to avoid confounding factors related to serious reproductive issues. Records with unclear diagnostic codes or incomplete cycle information were also removed.

[0050] For continuous data, use the Student's test (for categorical variables expressed as frequencies (percentages), and use the χ² test or Fisher's exact test to compare differences between groups) to exclude error data.

[0051] S2. Using the Light GBM model (1000 decision trees), samples were screened from embryonic morphological parameters and patient baseline indicators, respectively (ranked in descending order of feature importance (Gain score)). (See [link to relevant documentation]). Figure 1 , Figure 2 , Figure 3 (Selecting factors with a total importance of at least 70% as key factors) identified key factors influencing core clinical parameters (female BMI, baseline PRL, baseline T, baseline E2, AFC, total Gn medication use, daily HCG P, baseline FSH, daily HCG LH, baseline AMH, daily HCG E2; see [link]). Figure 4 Then, a Spear correlation analysis was performed between the key factors and each clinical core parameter (|ρ|≥0.3 and p<0.05). Key factors with significant correlations were designated as core factors (AFC, total Gn dosage, HCG daily P, baseline FSH, HCG daily LH, baseline AMH, HCG daily E2; see [link to relevant documentation]). Figure 5 , Figure 6 , Figure 7 );

[0052] S3. Reintegrate the core factors of each clinical core parameter into the light GBM model to construct a predictive model for each clinical core parameter (see [link]). Figure 8 , Figure 9 , Figure 10 );

[0053] S4. Evaluate the performance of the prediction model for each clinical core parameter, and optimize the prediction model based on the evaluation results to obtain a prediction model for embryo quality outcomes of women of childbearing age undergoing assisted reproductive technology.

[0054] The evaluation methods include: (1) drawing a scatter plot of the predicted values ​​and the actual values, and overlaying a 1:1 reference line and a regression fitting line;

[0055] The scatter plots included: the number of available embryos - the scatter plot showed a strong linear correlation between predicted and actual values ​​(Pearson correlation coefficient = 0.98, p < 0.001). The R² value was 0.962, indicating that the model could explain approximately 96.2% of the variance, and the prediction results had good explanatory power.

[0056] The 2PN total-scatter plot shows a strong linear correlation between predicted and actual values ​​(Pearson correlation coefficient = 0.98, p < 0.001). The R² value is 0.964, indicating that the model can explain approximately 96.4% of the variance variation, and the prediction results have good explanatory power.

[0057] The scatter plot of the number of high-quality embryos showed a strong linear correlation between the predicted and actual values ​​(Pearson correlation coefficient = 0.98, p < 0.001). The R² value was 0.975, indicating that the model could explain approximately 97% of the variance, and the prediction results had good explanatory power.

[0058] (2) Calculate the root mean square error (RMSE): the RMSE of the number of available embryos - scatter plot is 1.16; the RMSE of the total number of 2PNs - scatter plot is 1.231; the RMSE of the number of excellent embryos - scatter plot is 1.05.

[0059] Example 2

[0060] An application of a predictive model for embryo quality outcomes in women of reproductive age undergoing assisted reproductive technology (ART) is presented. The application method includes: visualizing and interpreting the embryo quality outcomes of women of reproductive age undergoing ART using the predictive model; specifically, SHAP analysis quantifies the global contribution of each feature to embryo quality outcomes through Summary Plot, Waterfall Plot reveals the driving path of single-sample prediction, and clinical thresholds are used to verify the nonlinear effects of key factors.

[0061] In the description of this specification, specific features, structures, materials, or characteristics may be combined in any suitable manner in one or more embodiments or examples.

[0062] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for constructing a predictive model for embryo quality outcomes in women of reproductive age undergoing assisted reproductive technology, characterized in that, Includes the following steps: S1. Obtain clinically relevant data from the sample; Statistical analysis was performed on the clinically relevant data of the sample to obtain the statistical data of the sample; The statistical analysis includes: continuous data are expressed as mean values, and categorical variables are expressed as frequencies; the clinically relevant data include: embryonic morphological parameters, patient baseline indicators, and core clinical parameters. S2. Using the Light GBM model, key factors affecting the core clinical parameters were screened from the embryonic morphological parameters and patient baseline indicators of the samples. Spear correlation analysis was then performed between the key factors and each core clinical parameter, and the key factors with significant correlation were identified as the core factors of each core clinical parameter. S3. The core factors of each clinical core parameter are re-incorporated into the light GBM model to construct a prediction model for each clinical core parameter. S4. Evaluate the performance of each clinical core parameter and optimize the prediction model based on the evaluation results to obtain a prediction model for embryo quality outcomes of women of childbearing age undergoing assisted reproductive technology.

2. The construction method according to claim 1, characterized in that, In S1, the embryo morphological parameters include: cell number, cell uniformity, cell fragmentation rate, embryo development speed at the corresponding date, and multinucleation phenomenon; the patient's basic indicators include: AFC, BMI, baseline PRL, baseline T, baseline E2, total Gn medication dosage, HCG daily P, baseline FSH, HCG daily LH, baseline AMH, and HCG daily E2 for women of childbearing age; the clinical core parameters include the number of excellent embryos, the number of usable embryos, and the total number of 2PN embryos.

3. The construction method according to claim 1, characterized in that, In S2, the Light GBM model includes at least 1000 decision trees.

4. The construction method according to claim 1, characterized in that, In S2, the screening method for the core factors includes: sorting the clinically relevant data in descending order of feature importance, and selecting the clinically relevant data with a total feature importance of not less than 70% as core factors.

5. The construction method according to claim 1, characterized in that, In S4, the evaluation method includes: (1) drawing a scatter plot of the predicted value and the true value, and overlaying a 1:1 reference line and a regression fitting line; (2) calculating the root mean square error RMSE.

6. The construction method according to claim 5, characterized in that, The scatter plots include: a scatter plot of available embryos, a scatter plot of total 2PNs, and a scatter plot of good embryos.

7. The construction method according to claim 6, characterized in that, In S4, the criteria for verifying that the prediction model meets the requirements are: the R² value of the scatter plot is ≥0.9; the RMSE of the scatter plot is ≥1.

8. A predictive model for embryo quality outcomes in women of reproductive age undergoing assisted reproductive technology, characterized in that, The prediction model is constructed using the construction method described in any one of claims 1-7.

9. The application of a predictive model for embryo quality outcomes in women of reproductive age undergoing assisted reproductive technology as described in claim 8, characterized in that, Used to predict embryo quality outcomes in women of reproductive age undergoing assisted reproductive technology.

10. The application according to claim 9, characterized in that, Application methods include: using predictive models to visualize and interpret embryo quality outcomes of women of reproductive age undergoing assisted reproductive technology.

Images