An integrated evaluation system for ovarian reserve, endocrine age and aging trend for female family planning

Abstract

The application relates to a female fertility planning-oriented ovary reserve, endocrine age and aging trend integrated evaluation system, which comprises the following modules: a data acquisition module for acquiring the age and anti-Mullerian hormone (AMH) level of a subject; a module for calculating the low ovarian response function, which is used for calculating the low ovarian response probability (p) of the subject by calculating the acquired data in the data acquisition module, and further converting the low ovarian response probability into a standardized ovary reserve score; a module for calculating the endocrine age of the subject, which is used for calculating the endocrine age of the subject by using the low ovarian response probability of the subject; and a module for predicting the age at which the ovary reserve of the subject changes, which is used for predicting the age or time limit at which the ovary reserve of the subject changes by using the endocrine age of the subject. The system and method can help women of childbearing age understand their best fertility time.

Description

Technical Field

[0001] This application relates to a system for optimizing the assessment of a subject's ovarian reserve function. This system can assess a subject's own ovarian reserve function and, based on the subject's current ovarian reserve status, estimate the number of years required for the subject to reach a specific ovarian reserve state, such as near-depletion of ovarian reserve, early decline in fertility due to ovarian reserve decline, or a significant decline in fertility due to ovarian reserve decline (or predict the age at which different individuals will reach the above-mentioned ovarian reserve state). Background Technology

[0002] The number of primordial follicles contained in the ovarian cortex is called ovarian reserve. It reflects the ovary's ability to provide healthy, fertile eggs and is the most important indicator of female ovarian function. Generally speaking, the more primordial follicles there are, the better their quality, and the higher the chance of conception.

[0003] However, many women of reproductive age are unaware that ovarian reserve varies significantly from person to person, referring to the number of primordial follicles in the ovarian cortex, ranging from tens of thousands to millions at birth. In our clinical practice, we have observed that some women maintain good ovarian reserve in their forties, while others face diminished ovarian reserve (DOR) or even depletion in their twenties. To improve infertility and achieve successful pregnancy, an increasing number of couples are seeking assisted reproductive technology (ART). However, not all couples benefit from ART. Its effectiveness is very limited in DOR patients or perimenopausal women, as their fertility has significantly declined or is nearing depletion, with few or no follicles in their ovaries. Internationally, there is a consensus that even expensive ovulation induction treatments cannot improve fertility in these individuals.

[0004] Although the number and quality of follicles change profoundly with age, the process of ovarian aging has not received enough attention. Women often only realize their fertility may be declining when they experience irregular menstruation or menopause, but by then their fertility is already extremely low and cannot be improved through assisted reproductive technologies. Early fertility decline due to diminished ovarian reserve usually occurs early, but the long-standing lack of clear methods for assessing ovarian reserve has caused many women of reproductive age to miss their optimal reproductive window. Summary of the Invention

[0005] As mentioned above, assessing a subject's current ovarian reserve is crucial for clinicians. By evaluating ovarian reserve, ovarian responsiveness, a vital clinical outcome in ovulation induction treatment, can be predicted. Previously, clinicians often relied on their own experience, judging based on factors such as age, body mass index, endocrine factor levels, and antral follicle count, which introduced a degree of subjectivity. Our system can accurately assess the ovarian reserve of subjects about to undergo treatment, assisting clinicians in developing more targeted treatment plans.

[0006] In summary, it is known that ovarian responsiveness is determined by ovarian reserve function. However, the inventors of this application take a reverse approach, using expected ovarian responsiveness to assess ovarian reserve function. The inventors first obtain the expected probability of low ovarian response based on the patient's basic information. Then, based on the system's pre-stored default ovarian reserve function grouping parameters, they group the subject's current ovarian reserve function to determine its level and assess the ovarian reserve level or ovarian youthfulness level.

[0007] Using the system developed by the inventors to assess the ovarian reserve function of a subject, the probability (p) of a subject’s low ovarian response can be calculated, that is, the subject’s current probability of low ovarian response, or the subject’s current ovarian reserve status.

[0008] Furthermore, the 'Fixed Interval' hypothesis exists internationally, suggesting a fixed time interval relationship between different ovarian reserve states. In other words, the rate of follicle depletion is roughly stable across the population; for example, the time interval between the early decline in ovarian reserve, the significant decline, and the eventual depletion of ovarian reserve is roughly the same for different individuals. This hypothesis is primarily based on the fact that the length of the menstrual cycle is generally stable, typically around 28 days. Based on this hypothesis, in this application, the inventors therefore speculate that the shape of the curve depicting the increase in the time interval of ovarian reserve with age is relatively fixed.

[0009] Based on the assumption that the rate of change in ovarian reserve with age is constant across the entire reproductive population, the inventors of this application attempt to predict, based on the current ovarian reserve status and the rate of ovarian reserve depletion (the rate of DOR with age), the time when a specific woman will experience early decline in ovarian reserve (25% probability of low response), the time when this decline will lead to a significant decrease in fertility (50% probability of low response), or the time when ovarian reserve is nearly exhausted, leading to near-exhaustion of fertility (95% probability of low response). The aim is to develop a method and system to help women predict the time (or age) to reach their desired ovarian reserve status based on their current ovarian reserve, which is significant for women's rational planning of their fertility and could be an effective way to reduce infertility rates among women of reproductive age.

[0010] Specifically, this application involves the following: 1. A system for predicting ovarian aging in a subject, comprising: The data acquisition module is used to acquire data on the age and anti-Müllerian hormone (AMH) levels of the subjects; The module for calculating ovarian low-response function is used to calculate the above data acquired in the data acquisition module, thereby calculating the probability (p) of ovarian low-response of the subject, and further converting the probability of ovarian low-response into a standardized ovarian reserve score. A module for calculating a subject's endocrine age, which uses the probability of a subject's low ovarian response to calculate the subject's endocrine age; and The module predicts the age at which changes in a subject's ovarian reserve occur, using the subject's endocrine age to predict the time or age at which new changes in ovarian reserve will occur.

[0011] 2. The system according to item 1, wherein, For women with regular menstrual cycles, the anti-Müllerian hormone (AMH) level refers to the concentration of anti-Müllerian hormone in the venous blood of the female subject on any given day. For women with irregular menstruation, the anti-Müllerian hormone (AMH) level refers to the highest concentration of anti-Müllerian hormone in venous blood during multiple consecutive venous blood draws in a female subject's menstrual cycle.

[0012] 3. The system according to item 1, wherein, Low ovarian response refers to a subject retrieving fewer than 5 eggs.

[0013] 4. The system according to item 1, wherein, In the module for calculating ovarian hyporesponsiveness, the probability (p) of ovarian hyporesponsiveness of a subject was calculated based on a logistic regression model using the subject's age and anti-Müllerian hormone (AMH) level.

[0014] 5. The system according to item 1, wherein, In the module for calculating low ovarian response function, the ovarian reserve score is calculated using the following conversion formula: Ovarian reserve score = (1 - probability of low ovarian response) × 100.

[0015] 6. The system according to item 1, wherein, In the module for calculating the subject's endocrine age, the subject's endocrine age is calculated using the probability of low ovarian response, based on any one of the following: a two-parameter logistic regression model, a two-parameter Probit model, a three-parameter Gompertz model, or a Weibul curve.

[0016] 7. The system according to item 6, wherein, The threshold is set to 0.15 in two-parameter logistic regression models, two-parameter Probit models, three-parameter Gompertz models, or Weibul curves.

[0017] 8. The system according to item 1, wherein the calculated probability of low ovarian response and standardized ovarian reserve score of the subject includes the following five cases: An ovarian low response rate of less than 5% is converted to an ovarian reserve score of greater than 95. An ovarian reserve score of 5% or higher and less than 25% is converted to an ovarian reserve score of 75 or higher and less than or equal to 95. An ovarian reserve score of 25% to 75% or less is equivalent to an ovarian reserve score of 25% to 50%. An ovarian reserve score of 50% to 50% with a low ovarian response rate is converted to an ovarian reserve score of 25 to 50. An ovarian reserve score of 75% to 25% is equivalent to an ovarian reserve score of 75% to 84%.

[0018] 9. The system according to item 4, wherein, In the module for calculating ovarian hyporesponsiveness, the probability (p) of ovarian hyporesponsiveness is calculated by performing a quadratic transformation on the subject's age and a cubic transformation on the subject's anti-Müllerian hormone (AMH) level data.

[0019] 10. The system according to item 9, wherein, Calculate the probability of low ovarian response in the subject (p The formula for () is as follows: Formula 1: (Formula 1) in, p The calculated probability of low ovarian response is used to characterize the subject's current ovarian reserve function. e represents the natural constant; Where a, b, c, d, f, and g are unitless parameters; age represents the subject's age, and AMH represents the subject's anti-Müllerian hormone (AMH) level; In the module for calculating ovarian reserve function, the subject's age and anti-Müllerian hormone (AMH) level are logarithmic and substituted into Formula 1 for calculation.

[0020] 11. The system according to item 10, wherein, 'a' is any value selected from -3.830 to -2.929. b is any value selected from 0.042 to 0.069. c is any value selected from 0.0017 to 0.0050. d is any value selected from -1.490 to -1.327. f is any value selected from 0.250 to 0.373. g is any value selected from 0.091 to 0.130.

[0021] 12. The system according to item 5, wherein, In the module for calculating the subject's endocrine age, based on the probability of low ovarian response, the endocrine age (age) is calculated by back-calculating from a two-parameter logistic curve using Formula 2 as shown below. Endocrine ): age Endocrine = (Formula 2) in, p The probability of low ovarian response for ovarian reserve function; a1 and b1 are estimates of the inflection point and growth rate parameters obtained based on the logistic curve, and are unitless parameters; Where a1 is a fixed value, preferably 37.553; b1 is a fixed value, preferably 0.28. 13. The system according to item 12, wherein, In the module for calculating the subject's endocrine age, the standard error of the endocrine age is calculated ( se age This allows for the calculation of endocrine age.Endocrine The 95% confidence interval of ) se age = in, p a2 is the calculated probability of low ovarian response, used to characterize the subject's current ovarian reserve function; b2 and c2 are unitless parameters. Where a2 is a fixed value of 0.283; b2 is a fixed value of 0.012; and c2 is a fixed value of 0.168. The lower and upper limits of the 95% confidence interval for calculating endocrine age are as follows: age Endocrine_lower = age Endocrine -1.96 se age age Endocrine_upper = age Endocrine +1.96 se age .

[0022] 14. The system according to item 1, wherein, In the module predicting the age at which a subject's ovarian reserve changes, the following formula (Formula 3) is used to calculate the number of years from the subject's current ovarian reserve status to the point where a significant decline in ovarian reserve leads to a significant decline in fertility, i.e., the number of years when the probability of low ovarian response is 50%: Based on endocrine age, the number of years until ovarian reserve reaches 50 points (D1) is calculated, which is obtained from the logistic curve. p =0.5 times the difference between age and endocrine age; D1=age( )-age Endocrine =a1 - age Endocrine (Formula 3) p The probability of low ovarian response for ovarian reserve function; a1 and b1 are estimates of the inflection point and growth rate parameters obtained based on the logistic curve, and are unitless parameters; Preferably, a1 can be any value from 37.224 to 37.883, and more preferably, a1 is 37.553; b1 can be any value from 0.260 to 0.301, and more preferably, b1 is 0.283; age Endocrine Any value between the lower and upper limits of the 95% confidence interval in Formula 2 can be used to further optimize age.Endocrine This is the calculated value from Formula 2.

[0023] 15. The system according to item 14, wherein the age at which ovarian reserve reaches 50 points is calculated using the following formula. DOR ): age DOR = age chronological + D1 Among them, age chronological This represents the subject's current age, and D1 is the D1 value calculated using Formula 3.

[0024] 16. The system according to item 1, wherein, In the module predicting the age at which a subject's ovarian reserve changes, the following Formula 4 is used to calculate the number of years from the subject's current ovarian reserve to the onset of perimenopause, i.e., the number of years when the probability of low ovarian response is 95%: Based on endocrine age, the number of years to perimenopause (D2) is calculated, which is obtained from the logistic curve. p The difference between age and endocrine age at a value of 0.95: D2=age( )-age Endocrine =a1 - age Endocrine (Formula 4) p The probability of low ovarian response for ovarian reserve function; a1 and b1 are estimates of the inflection point and growth rate parameters obtained based on the logistic curve, and are unitless parameters; a1 can be any value from 37.224 to 37.883, with a1 being more preferably 37.553; b1 can be any value from 0.260 to 0.301, with b1 being more preferably 0.283; age Endocrine Any value between the lower and upper limits of the 95% confidence interval in Formula 2 can be used, further aging. Endocrine The calculated value of Formula 2 is preferred.

[0025] 17. The system according to item 16, wherein the age at onset of perimenopause is calculated using the following formula. 围绝经 ): age 围绝经 = age chronological + D2 Among them, age chronological D represents the subject's current age, and D2 is the D2 value calculated using Formula 4.

[0026] 18. A method for predicting ovarian aging in a subject, comprising: The data acquisition steps involve obtaining data on the age and anti-Müllerian hormone (AMH) levels of the subjects. The step of calculating low ovarian response function involves calculating the data obtained in the data acquisition step to calculate the probability (p) of low ovarian response of the subject, and further converting the probability of low ovarian response into a standardized ovarian reserve score. The steps for calculating a subject's endocrine age utilize the subject's low ovarian response probability to determine their endocrine age; and The steps for predicting the age at which changes in a subject's ovarian reserve occur utilize the subject's endocrine age to predict the time or age at which new changes in ovarian reserve will occur.

[0027] 19. The method according to item 18, wherein, For women with regular menstrual cycles, the anti-Müllerian hormone (AMH) level refers to the concentration of anti-Müllerian hormone in the venous blood of the female subject on any given day. For women with irregular menstruation, the anti-Müllerian hormone (AMH) level refers to the highest concentration of anti-Müllerian hormone in venous blood during multiple consecutive venous blood draws in a female subject's menstrual cycle.

[0028] 20. The method according to item 18, wherein, Low ovarian response refers to a subject retrieving fewer than 5 eggs.

[0029] 21. The method according to item 18, wherein, In the step of calculating ovarian hyporesponsiveness, the probability (p) of ovarian hyporesponsiveness of the subject was calculated based on the subject's age and anti-Müllerian hormone (AMH) level using a logistic regression model.

[0030] 22. The method according to item 18, wherein, In calculating low ovarian response function, the ovarian reserve score is calculated using the following conversion formula: Ovarian reserve score = (1 - probability of low ovarian response) × 100.

[0031] 23. The method according to item 18, wherein, In the step of calculating the subject's endocrine age, the subject's endocrine age is calculated using the probability of low ovarian response, based on any one of the following: a two-parameter logistic regression model, a two-parameter Probit model, a three-parameter Gompertz model, or a Weibul curve.

[0032] 24. The method according to item 23, wherein, The threshold is set to 0.15 in two-parameter logistic regression models, two-parameter Probit models, three-parameter Gompertz models, or Weibul curves.

[0033] 25. The method according to item 18, wherein the calculated probability of low ovarian response and standardized ovarian reserve score of the subject includes the following five categories: An ovarian low response rate of less than 5% is converted to an ovarian reserve score of greater than 95. An ovarian reserve score of 5% or higher and less than 25% is converted to an ovarian reserve score of 75 or higher and less than or equal to 95. An ovarian reserve score of 25% to 75% or less is equivalent to an ovarian reserve score of 25% to 50%. An ovarian reserve score of 50% to 50% with a low ovarian response rate is converted to an ovarian reserve score of 25 to 50. An ovarian reserve score of 75% to 25% is equivalent to an ovarian reserve score of 75% to 84%.

[0034] 26. The method according to item 21, wherein, In the step of calculating ovarian hyporesponsiveness, the probability (p) of ovarian hyporesponsiveness of the subject was calculated by performing a quadratic transformation on the subject's age and a cubic transformation on the subject's anti-Müllerian hormone (AMH) level data.

[0035] 27. The method according to item 26, wherein, Calculate the probability of low ovarian response in the subject ( p The formula for () is as follows: Formula 1: (Formula 1) in, p The calculated probability of low ovarian response is used to characterize the subject's current ovarian reserve function. e represents the natural constant; Where a, b, c, d, f, and g are unitless parameters; age represents the subject's age, and AMH represents the subject's anti-Müllerian hormone (AMH) level; In the step of calculating ovarian reserve function, the subject's age and anti-Müllerian hormone (AMH) level are logarithmic and substituted into Formula 1 for calculation.

[0036] 28. The method according to item 27, wherein, 'a' is any value selected from -3.830 to -2.929. b is any value selected from 0.042 to 0.069. c is any value selected from 0.0017 to 0.0050. d is any value selected from -1.490 to -1.327. f is any value selected from 0.250 to 0.373. g is any value selected from 0.091 to 0.130.

[0037] 29. The method according to item 22, wherein, In the step of calculating the subject's endocrine age, based on the probability of low ovarian response, the endocrine age (age) is calculated by back-calculating from a two-parameter logistic curve using the following formula (Formula 2). Endocrine ): age Endocrine = (Formula 2) in, p The probability of low ovarian response for ovarian reserve function; a1 and b1 are estimates of the inflection point and growth rate parameters obtained based on the logistic curve, and are unitless parameters; Where a1 is a fixed value, preferably 37.553; b1 is a fixed value, preferably 0.28. 30. The method according to item 29, wherein, In the step of calculating the subject's endocrine age, the standard error of the endocrine age is calculated. se age This allows for the calculation of endocrine age. Endocrine The 95% confidence interval of ) se age = in, p a2 is the calculated probability of low ovarian response, used to characterize the subject's current ovarian reserve function; b2 and c2 are unitless parameters. Where a2 is a fixed value of 0.283; b2 is a fixed value of 0.012; and c2 is a fixed value of 0.168. The lower and upper limits of the 95% confidence interval for calculating endocrine age are as follows: age Endocrine_lower = age Endocrine -1.96 se age age Endocrine_upper = age Endocrine +1.96 se age .

[0038] 31. The method according to item 18, wherein, In the step of predicting the age at which a subject's ovarian reserve changes, the following formula (Formula 3) is used to calculate the number of years from the subject's current ovarian reserve status to the point where a significant decline in ovarian reserve leads to a significant decline in fertility, i.e., the number of years at which the probability of low ovarian response is 50%: Based on endocrine age, the number of years until ovarian reserve reaches 50 points (D1) is calculated, which is obtained from the logistic curve. p =0.5 times the difference between age and endocrine age; D1=age( )-age Endocrine =a1 - age Endocrine (Formula 3) p The probability of low ovarian response for ovarian reserve function; a1 and b1 are estimates of the inflection point and growth rate parameters obtained based on the logistic curve, and are unitless parameters; Preferably, a1 can be any value from 37.224 to 37.883, and more preferably, a1 is 37.553; b1 can be any value from 0.260 to 0.301, and more preferably, b1 is 0.283; age Endocrine Any value between the lower and upper limits of the 95% confidence interval in Formula 2 can be used to further optimize age. Endocrine This is the calculated value from Formula 2.

[0039] 32. According to the method described in item 31, the age at which ovarian reserve reaches 50 points is calculated using the following formula. DOR ): age DOR = age chronological + D1 Among them, age chronological This represents the subject's current age, and D1 is the D1 value calculated using Formula 3.

[0040] 33. The method according to item 18, wherein, In the step of predicting the age at which a subject's ovarian reserve changes, the following formula (Formula 4) is used to calculate the number of years from the subject's current ovarian reserve to the onset of perimenopause, i.e., the number of years when the probability of low ovarian response is 95%: Based on endocrine age, the number of years until perimenopause (D2) is calculated, which is obtained from the logistic curve. p The difference between age and endocrine age at a value of 0.95: D2=age( )-age Endocrine =a1 - age Endocrine (Formula 4) p The probability of low ovarian response for ovarian reserve function; a1 and b1 are estimates of the inflection point and growth rate parameters obtained based on the logistic curve, and are unitless parameters; a1 can be any value from 37.224 to 37.883, with a1 being more preferably 37.553; b1 can be any value from 0.260 to 0.301, with b1 being more preferably 0.283; age Endocrine Any value between the lower and upper limits of the 95% confidence interval in Formula 2 can be used, further aging. Endocrine The calculated value of Formula 2 is preferred.

[0041] 34. The method according to item 33, wherein the age at onset of perimenopause is calculated using the following formula (age). 围绝经 ): age 围绝经 = age chronological + D2 Among them, age chronological D represents the subject's current age, and D2 is the D2 value calculated using Formula 4.

[0042] Invention Effects The system described in this application can identify individuals with diminished ovarian reserve (DOR) at an early stage. Using the system and method described in this application, participants can predict the timeframe at which key changes in their ovarian reserve will occur, and be advised to attempt conception as early as possible before entering DOR. While the quantity and quality of follicles change profoundly with age, the process of ovarian aging has not received sufficient attention. People often only realize their fertility may be declining when menstrual irregularities or menopause occur, but by then, fertility is already extremely low and cannot be improved through assisted reproductive technologies. Early fertility decline due to diminished ovarian reserve usually occurs early, but due to the long-standing lack of clear methods for assessing ovarian reserve, many women of reproductive age miss their optimal reproductive window. However, as mentioned above, the system and method described in this application can accurately help participants predict when ovarian reserve begins to decline, when it becomes significantly diminished, or when it approaches depletion. Such a system and method can help women of reproductive age understand their optimal reproductive window in a timely manner. Attached Figure Description

[0043] Various other advantages and benefits of this application will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiments below. The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of this application. It is obvious that the drawings described below are merely some embodiments of this application, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort. Furthermore, the same reference numerals denote the same parts throughout the drawings.

[0044] Picture 1 The calibration of each model is shown. (A) is the training queue (2017-2019); (B) is the testing queue (2020-2021). The top plot covers the entire 0-1.0 range; the bottom plot is scaled to 0-0.25. The black dashed line equals the ideal. Model-0 has an overestimation risk <0.60; Model-1 improves fit; Model-2 is closest to the ideal line.

[0045] Picture 2This diagram illustrates the trajectory of age-related decline in female fertility. (A) A conceptual illustration shows four key milestones—low fertility, infertility, menstrual irregularities, and menopause—modeled as an S-shaped function of age. The 50% probability lines intersect these curves at approximately ages 31, 41, 46, and 51, indicating that reproductive events gradually change with age. (B) Empirical ovarian aging curves from 31,923 first-cycle ART patients: the observed proportion of reduced ovarian reserve (DOR, black dots) versus age, fitted using a two-parameter logistic regression model. High consistency confirms an S-shaped increase in DOR risk, which is... Picture 2 A. The basis of the diagram.

[0046] Picture 3 Displays the dynamics and diagnostic performance of AMH during the cycle. (A) Serum AMH distributions measured on days 2, 6, and 12 of the cycle (first row) are highly right-skewed; logarithmic transformation (second row) normalizes the data, as shown in the histogram, density curve, and accompanying box-and-whisker plot. (B) The least-squares (LS) mean of Ln(AMH) gradually decreases throughout the follicular phase (day 2 > day 6 > day 12); error bars indicate 95% CI, and an inserted table lists paired differences. (C) ROC curves comparing POR prediction models for AMH samples taken on days 2, 6, or 12. Detailed Implementation

[0047] Specific embodiments of this application will now be described in more detail. However, it should be understood that this application can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of this application and to fully convey the scope of this application to those skilled in the art.

[0048] It should be noted that certain terms are used in the specification and claims to refer to specific components. Those skilled in the art will understand that different terms may be used to refer to the same component. This specification and claims do not distinguish components based on differences in terminology, but rather on differences in function. The terms "comprising" or "including" used throughout the specification and claims are open-ended and should be interpreted as "comprising but not limited to." The following descriptions in the specification are preferred embodiments for carrying out this application; however, these descriptions are for the purpose of understanding the general principles of the specification and are not intended to limit the scope of this application. The scope of protection of this application shall be determined by the appended claims.

[0049] In this application, ovarian reserve refers to the number of primordial follicles contained in the ovarian cortex, known as ovarian reserve. It reflects the ovary's ability to provide healthy, fertile eggs and is the most important indicator of female ovarian function. Generally speaking, the more primordial follicles there are, the better their quality, and the higher the chance of conception.

[0050] In this application, low ovarian response, also known as decreased ovarian reserve (DOR), refers to the number of oocytes retrieved on the day of oocyte retrieval in a reproductive cycle being less than 5 (i.e., 0-4).

[0051] However, the number of primordial follicles cannot be non-invasively assessed; it can only be assessed by the number of follicles mobilized in each menstrual cycle. If too few follicles are mobilized in an IVF-ET cycle (low ovarian response), it indicates a decline in ovarian reserve function, i.e., ovarian fertility.

[0052] Age is generally considered the most important factor in evaluating ovarian reserve. A study on age and IVF success rate showed that the IVF success rate for women under 30 was about 26%, while the success rate for women aged 37 and above was only 9%.

[0053] The mechanisms by which ovarian reserve declines with age are as follows: (I) Decreased number of follicles: Primordial follicles appear after embryonic sex differentiation, at which time the number of follicles is at its maximum. After puberty, follicles begin to mature, and with the completion of ovulation, a large number of follicles that are recruited but not released atrophy and disappear to form the corpus luteum. The number of follicles decreases continuously with age: the human embryo at 20 weeks of age has the most, about 6 million follicles, which decreases to 700,000-2 million in the neonatal period, about 40,000 in puberty, and only about a thousand at the beginning of menopause, until it is completely depleted. (II) Decreased egg quality: Embryo quality is mainly determined by egg quality. Older age can lead to an increased probability of aneuploidy in oocytes, an increased risk of mitochondrial dysfunction, loss of oocyte polarity, and epigenetic changes in oocytes. (III) Endocrine factors: The hypothalamus-pituitary-ovarian axis regulates the female menstrual cycle and ovulation. Abnormal endocrine levels in this axis can lead to infertility. AMH and inhibin B are secreted by small follicles and are a direct reflection of ovarian reserve. As women age, ovarian reserve decreases, leading to a reduction in the number of recruitable follicles and consequently, a decrease in the concentrations of AMH and inhibin B secreted by the follicles. Inhibin B can negatively regulate pituitary FSH secretion; a decrease in inhibin B levels results in increased FSH secretion during the luteal phase. This premature increase in FSH promotes the growth of new follicles and E2 secretion, ultimately shortening the menstrual cycle. Increased serum FSH levels and decreased inhibin B levels indicate reduced follicle sensitivity to FSH, suggesting a decrease in the number of recruitable antral follicles. The menstrual cycle reflects ovarian reserve and fertility; advanced ovarian age leads to a shorter menstrual cycle. A reduction of 2-3 days in the menstrual cycle is a sensitive indicator of reproductive system aging, suggesting premature follicle growth (elevated FSH levels) and a decline in primordial follicle reserve, i.e., youthfulness.

[0054] To address the problems existing in the prior art, this application provides a system for predicting ovarian aging in a subject, comprising: The data acquisition module is used to acquire data on the age and anti-Müllerian hormone (AMH) levels of the subjects; The module for calculating ovarian low-response function is used to calculate the above data acquired in the data acquisition module, thereby calculating the probability (p) of ovarian low-response of the subject, and further converting the probability of ovarian low-response into a standardized ovarian reserve score. A module for calculating a subject's endocrine age, which uses the probability of a subject's low ovarian response to calculate the subject's endocrine age; and The module predicts the age at which changes in a subject's ovarian reserve occur, using the subject's endocrine age to predict the time or age at which new changes in ovarian reserve will occur.

[0055] In the data acquisition module, different methods were used to measure anti-Müllerian hormone (AMH) levels depending on whether the subject's menstrual cycle was regular. If the menstrual cycle was irregular, AMH levels may fluctuate significantly within the same cycle and between different cycles, so relying on a single test may not accurately reflect the true ovarian reserve.

[0056] For women with regular menstrual cycles, the anti-Müllerian hormone (AMH) level refers to the concentration of anti-Müllerian hormone in the venous blood of the female subject on any given day; for women with irregular menstrual cycles, the anti-Müllerian hormone (AMH) level refers to the highest concentration of anti-Müllerian hormone in the venous blood of the female subject during multiple consecutive venous blood draws in her menstrual cycle.

[0057] In this application, regular menstruation means a menstrual cycle of 21-35 days, each menstrual period lasting 3 to 7 days, with a fixed amount of bleeding and menstrual symptoms; irregular menstruation means a menstrual cycle shorter than 21 days or longer than 35 days, or a menstrual interval that is not constant.

[0058] In some implementations, the input range for AMH concentration values ​​is 0.06–41.4 ng / ml.

[0059] In the module for calculating ovarian hyporesponsiveness, a logistic regression model was used to calculate the probability (p) of ovarian hyporesponsiveness based on the subject's age and anti-Müllerian hormone (AMH) level. Specifically, the probability (p) of ovarian hyporesponsiveness was calculated by performing a quadratic transformation on the subject's age and a cubic transformation on the subject's AMH level data.

[0060] In some implementations, the probability of a subject having a low ovarian response is calculated. p The formula for () is as follows: Formula 1: (Formula 1) in, p The calculated probability of low ovarian response is used to characterize the subject's current ovarian reserve function. e represents the natural constant; Where a, b, c, d, f, and g are unitless parameters; age represents the subject's age, and AMH represents the subject's anti-Müllerian hormone (AMH) level; In the module for calculating ovarian reserve function, the subject's age and anti-Müllerian hormone (AMH) level are logarithmic and substituted into Formula 1 for calculation.

[0061] In some implementations, a is any value selected from -3.830 to -2.929, b is any value selected from 0.042 to 0.069, c is any value selected from 0.0017 to 0.0050, d is any value selected from -1.490 to -1.327, f is any value selected from 0.250 to 0.373, and g is any value selected from 0.091 to 0.130.

[0062] In this application, the purpose of converting the probability of poor ovarian response into an ovarian reserve score is to transform the predicted risk of poor response into an intuitive scoring indicator. Originally, the probability of poor response reflected the risk of low oocyte retrieval during ovulation induction. After this conversion, a high score indicates low risk and good ovarian reserve; a low score indicates higher risk and poor ovarian function. Through this conversion, a lower probability of poor response corresponds to a higher score, intuitively demonstrating the quality of an individual's ovarian reserve. In this way, the scoring aligns better with clinical intuition, helping physicians quickly assess an individual's ovarian function status. It also facilitates integration with other reproductive indicators, providing a reliable basis for risk stratification and personalized treatment.

[0063] In some implementations, the ovarian reserve score is calculated using the following conversion formula: Ovarian reserve score = (1 - probability of low ovarian response) × 100.

[0064] In some implementations, the calculated probability of poor ovarian response and standardized ovarian reserve score for the subject includes the following five scenarios: An ovarian low response rate of less than 5% is converted to an ovarian reserve score of greater than 95. An ovarian reserve score of 5% or higher and less than 25% is converted to an ovarian reserve score of 75 or higher and less than or equal to 95. An ovarian reserve score of 25% to 75% or less is equivalent to an ovarian reserve score of 25% to 50%. An ovarian reserve score of 50% to 50% with a low ovarian response rate is converted to an ovarian reserve score of 25 to 50. An ovarian reserve score of 75% to 25% is equivalent to an ovarian reserve score of 75% to 84%.

[0065] In this application, the probability of low ovarian response is less than or equal to 5%; the ovarian reserve grade is A (ovarian reserve score ≥ 95 points). This reserve score indicates that the subject has very sufficient ovarian reserve, suggesting excellent fertility, suitability for natural pregnancy, and high flexibility in planned pregnancy. Good ovarian reserve helps maintain sex hormone balance and reduces the risk of cardiovascular disease and osteoporosis.

[0066] A low ovarian response rate greater than 5% and less than or equal to 25% indicates an ovarian reserve grade of B (ovarian reserve score of 75-<95). This score suggests that the subject has good ovarian reserve and relatively high fertility, but a slight downward trend has begun to emerge. Early planning for pregnancy is recommended.

[0067] A low ovarian response rate greater than 25% and less than or equal to 50% indicates ovarian reserve classification as Grade C (ovarian reserve score 50-<75). This score indicates moderate ovarian reserve and declining fertility. When ovarian reserve drops to 50, menopause may be approximately 10 years away, at which point the rate of decline begins to accelerate significantly, requiring close monitoring of fertility planning. Early signs of declining ovarian reserve may include menstrual cycle changes and mild metabolic disturbances, such as weight gain or altered fat distribution.

[0068] A low ovarian response rate greater than 50% and less than or equal to 75% indicates an ovarian reserve grade of D (ovarian reserve score 25-<50). This score suggests low ovarian reserve and a significantly reduced likelihood of natural pregnancy; assisted reproductive technology support is recommended as soon as possible. Decreased ovarian function may affect bone health, increase the risk of bone loss, and may cause mood swings and mild memory loss.

[0069] A low ovarian response rate greater than 75% indicates an ovarian reserve grade of E (ovarian reserve score <25). This score suggests extremely low ovarian reserve, near-loss of fertility, and potential imminent menopause. Special attention should be paid to overall health, as declining ovarian function can lead to a significant drop in estrogen levels, increasing the risk of atherosclerosis, osteoporosis, and sleep disorders.

[0070] When the system of this application is used as software, a five-color management system for ovarian reserve assessment can be constructed, thereby displaying the above-mentioned ovarian reserve classification through different colors. For example, green can represent grade A, light green represents grade B, yellow represents grade C, orange represents grade D, and red represents grade E. Different classifications require different monitoring frequencies, for example: grade A: ≥12 months for follow-up; grade B: 6–12 months for follow-up; grade C: 6 months for follow-up; grade D: 3–6 months for follow-up; grade E: ≤3 months or follow-up as recommended by experts.

[0071] Based on the five-color management system, participants should combine quantitative risk assessment with their marital partner's situation to do a good job in fertility reserve management, flexibly choose egg or embryo freezing, and achieve personalized fertility management of "early assessment, early intervention, and early preservation", providing a solid guarantee for individualized fertility and women's health management.

[0072] Ovarian endocrine age (OIA) is an indicator that maps an individual's ovarian reserve score to a population-wide ovarian aging curve. Its purpose is to reflect the relative state of an individual's ovarian function, rather than their chronological age. Specifically, by measuring anti-Müllerian hormone (AMH) levels and combining them with chronological age, a statistical model is used to predict the probability of a low response during ovulation induction. This probability is then standardized into an OIA score. Subsequently, an S-shaped ovarian aging curve is constructed using a large amount of population data to determine the trend of ovarian function at different ages. The age corresponding to the individual's score on the curve is then identified as the OIA endocrine age. If a young woman's OIA score corresponds to an endocrine age on the aging curve that is significantly higher than her chronological age, it may indicate a rapid decline in ovarian function and a potential premature decrease in future fertility. Conversely, if the chronological age and endocrine age are close to or below the average, it suggests that ovarian function is at a relatively ideal level.

[0073] In the module for calculating the subject's ovarian endocrine age, the subject's endocrine age is calculated using the probability of low ovarian response, based on any one of the following: a two-parameter logistic regression model, a two-parameter Probit model, a three-parameter Gompertz model, or a Weibul curve.

[0074] Specifically, this application generates a typical S-shaped curve based on any one of the following models: two-parameter logistic regression, two-parameter Probit, three-parameter Gompertz, or Weibull curve. By mapping the standardized ovarian reserve score of the subjects onto this S-shaped ovarian aging curve, the individual's ovarian endocrine age can be accurately calculated. This mapping process reflects the degree of matching between the individual's ovarian reserve level and the ovarian aging trend of the entire population, thus providing a strong basis for subsequent reproductive health assessment and individualized treatment plans. When constructing the ovarian aging curve, the inventors of this application map the individual's ovarian reserve score onto an S-shaped curve reflecting the aging trend of the population. This mapping relies on setting an appropriate threshold (i.e., the predicted probability of low ovarian response or ovarian reserve score) to determine whether there is a decline in ovarian reserve (DOR). Through comparative analysis of four models—two-parameter logistic regression, two-parameter Probit, three-parameter Gompertz, and Weibull growth curve—the inventors found that when the threshold is set to 0.15, various statistical indicators (such as AICc, BIC, SSE, MSE, RMSE, and R) are significantly improved. 2 All of these are better than the case where R is set to 0.5. 2 The increase in the value was particularly significant, and this threshold was highly consistent with the probability of low ovarian response in the actual population.

[0075] Specifically, after adopting a cutoff of 0.15, the fitting performance of both logistic regression and other models was significantly improved. For example, the logistic regression model constructed based on a cubic term model to predict the probability of low ovarian response (ovarian reserve score) showed a significantly improved R-value at a cutoff of 0.15. 2 The R-value of the Weibull model is as high as 0.9911. 2 It also reaches as high as 0.9930, while Rc is higher when cutoff = 0.5. 2 The values ​​are 0.9877 and 0.9873, indicating that the model is better when the tangent point is 0.15.

[0076] The threshold optimization process implemented in this application ensures that the model can more accurately capture subtle changes in future ovarian function when predicting the risk of low ovarian response and deducing ovarian endocrine age, thereby providing a more reliable basis for clinical decision-making. In this application, setting the threshold to 0.15 not only conforms to the principles of statistical optimization but also more closely reflects clinical reality, with significant results.

[0077] In some implementations, based on the probability of low ovarian response, endocrine age is calculated by back-calculating from a two-parameter logistic curve using Formula 2 as follows. Endocrine ): age Endocrine = (Formula 2) in, p The probability of low ovarian response for ovarian reserve function; a1 and b1 are estimates of the inflection point and growth rate parameters obtained based on the logistic curve, and are unitless parameters; Where a1 is a fixed value, preferably 37.553; b1 is a fixed value, preferably 0.283.

[0078] In some implementations, the standard error for calculating endocrine age is ( se age This allows for the calculation of endocrine age. Endocrine The 95% confidence interval of ) se age = in, p a2 is the calculated probability of low ovarian response, used to characterize the subject's current ovarian reserve function; b2 and c2 are unitless parameters. Where a2 is a fixed value of 0.283; b2 is a fixed value of 0.012; and c2 is a fixed value of 0.168. The lower and upper limits of the 95% confidence interval for calculating endocrine age are as follows: The standard error for calculating endocrine age ( se age This allows for the calculation of endocrine age. Endocrine The 95% confidence interval of ) se age = in, p a2, b2, and c2 are the calculated probability of low ovarian response, which characterizes the subject's current ovarian reserve function; these are unitless parameters. Where a2 is a fixed value of 0.283; b2 is a fixed value of 0.012; and c2 is a fixed value of 0.168. The lower and upper limits of the 95% confidence interval for calculating endocrine age are as follows: age Endocrine_lower = age Endocrine -1.96 se age age Endocrine_upper = age Endocrine +1.96 se age .

[0079] The significance of calculating the 95% confidence interval for endocrine age in this application lies in two aspects. First, it quantifies the uncertainty of the estimate. Ovarian endocrine age is a point estimate obtained by substituting the subject's ovarian reserve score into a two-parameter logistic curve. However, the model is derived from a finite sample, and the parameters contain errors. The 95% confidence interval (CI) takes these random errors into account: if the sampling is repeated and the model is rebuilt under the same conditions, approximately 95% of the intervals obtained will cover the "true" endocrine age. It visually demonstrates the accuracy of the estimate using the interval width—the narrower the interval, the more confident the model; the wider the interval, the more cautious the interpretation. Second, it distinguishes statistically significant differences. Comparing the subject's actual age with the 95% CI of the endocrine age can determine whether the difference is statistically significant. If the actual age falls outside the interval, it suggests that ovarian function deviates significantly from the average level for the same age; if it falls within the interval, the difference may only be caused by random errors. This helps in the clinical identification of true "premature" or "delayed" ovarian aging. Third, it supports individualized clinical decision-making. For women requiring intervention (such as early childbearing or egg freezing), the CI provides a confidence range for risk assessment: a narrow interval exceeding the actual age indicates a clear problem and urgent action; a wide interval or exceeding the actual age requires further testing or follow-up in conjunction with other indicators. This avoids both overconfidence leading to aggressive decisions and excessive caution causing delays. Finally, it promotes risk communication and continuous model optimization. Reporting point estimates and 95% CI simultaneously helps subjects and medical staff visually see "outcome + confidence," improving transparency. It also allows developers to monitor changes in interval width across different populations and laboratory batches, identifying potential model deficiencies in specific subgroups and iteratively optimizing the model. In short, the 95% confidence interval not only tells us "what the predicted value is," but also answers "how confident we are in this prediction," a key element in elevating the system from simple numerical output to an interpretable and reliable clinical tool.

[0080] In the module predicting the age at which a subject's ovarian reserve changes, the following formula (Formula 3) is used to calculate the number of years from the subject's current ovarian reserve status to the point where a significant decline in ovarian reserve leads to a significant decline in fertility, i.e., the number of years when the probability of low ovarian response is 50%: Based on endocrine age, the number of years until ovarian reserve reaches 50 points (D1) is calculated, which is obtained from the logistic curve. p =0.5 times the difference between age and endocrine age; D1=age( )-age Endocrine =a1 - age Endocrine (Formula 3) pThe probability of low ovarian response for ovarian reserve function; a1 and b1 are estimates of the inflection point and growth rate parameters obtained based on the logistic curve, and are unitless parameters; Preferably, a1 can be any value from 37.224 to 37.883, and more preferably, a1 is 37.553; b1 can be any value from 0.260 to 0.301, and more preferably, b1 is 0.283; age Endocrine Any value between the lower and upper limits of the 95% confidence interval in Formula 2 can be used to further optimize age. Endocrine This is the calculated value from Formula 2.

[0081] Furthermore, the age at which ovarian reserve reaches 50 points is calculated using the following formula. DOR ): age DOR = age chronological + D1 Among them, age chronological This represents the subject's current age, and D1 is the D1 value calculated using Formula 3.

[0082] In the module predicting the age at which a subject's ovarian reserve changes, the following Formula 4 is used to calculate the number of years from the subject's current ovarian reserve to the onset of perimenopause, i.e., the number of years when the probability of low ovarian response is 95%: Based on endocrine age, the number of years to perimenopause (D2) is calculated, which is obtained from the logistic curve. p The difference between age and endocrine age at a value of 0.95: D2=age( )-age Endocrine =a1 - age Endocrine (Formula 4) p The probability of low ovarian response for ovarian reserve function; a1 and b1 are estimates of the inflection point and growth rate parameters obtained based on the logistic curve, and are unitless parameters; a1 can be any value from 37.224 to 37.883, with a1 being more preferably 37.553; b1 can be any value from 0.260 to 0.301, with b1 being more preferably 0.283; age Endocrine Any value between the lower and upper limits of the 95% confidence interval in Formula 2 can be used, further aging. Endocrine The calculated value of Formula 2 is preferred.

[0083] Furthermore, the age at the onset of perimenopause can be calculated using the following formula: perimenopause = agechronological + D2 Where agechronological represents the subject's current age, and D2 is the D2 value calculated using Formula 4. This application also provides a method for predicting ovarian aging in a subject, comprising: The data acquisition steps involve obtaining data on the age and anti-Müllerian hormone (AMH) levels of the subjects. The step of calculating low ovarian response function involves calculating the data obtained in the data acquisition step to calculate the probability (p) of low ovarian response of the subject, and further converting the probability of low ovarian response into a standardized ovarian reserve score. The steps for calculating a subject's endocrine age utilize the subject's low ovarian response probability to determine their endocrine age; and The steps for predicting the age at which changes in a subject's ovarian reserve occur utilize the subject's endocrine age to predict the time or age at which new changes in ovarian reserve will occur.

[0084] The data acquisition steps, the steps for calculating low ovarian response function, the steps for calculating the subject's endocrine age, and the steps for predicting the age at which the subject's ovarian reserve changes are involved in the method, respectively, correspond to the data acquisition system, the system for calculating low ovarian response function, the system for calculating the subject's endocrine age, and the system for predicting the age at which the subject's ovarian reserve changes, respectively, involved in the above-mentioned system.

[0085] Example Example 1 Controlled ovarian stimulation (COS) therapy Gn (recombinant human FSH) treatment begins on day 2 or 3 of the menstrual cycle. The starting dose is selected based on age, BMI (Body Mass Index, calculated by dividing weight in kilograms by height in meters squared, a commonly used international standard for measuring body fat and health), and FSH and AFC levels on days 2-4 of the menstrual cycle. During ovulation induction, the starting Gn dose is adjusted based on ultrasound observation and serum E2 levels. GnRH antagonist treatment begins on days 5-7 of stimulation, when the developing follicles are 10-12 mm in diameter. When at least two dominant follicles (≥18 mm in diameter) are visible on ultrasound, 5000-10000 IU of hCG is administered to induce final oocyte maturation. Oocyte retrieval is performed 36 hours after hCG administration. 1-3 embryos are transferred or cryopreserved. Lutein phase progesterone support is then provided.

[0086] In the embodiments of this application, subjects who received the aforementioned GnRH antagonist treatment between 2017 and 2021 were utilized, and serum AMH levels, actual age, and relevant clinical variables were recorded at baseline. To ensure data integrity, stringent quality control measures were implemented, excluding records with missing values ​​or extreme outliers. Ultimately, 16,327 antagonist cycles performed between January 2017 and December 2019 constituted the training cohort, while 15,596 cycles performed between January 2020 and December 2021 constituted an independent validation cohort for constructing the system involved in this application.

[0087] Sample acquisition and endocrine assays For subjects as described above, venous blood samples were drawn and immediately inverted five times to promote thorough blood clotting. Serum was collected by centrifugation and used for endocrine assessment. Serum AMH concentrations in subjects were measured using an ultrasensitive two-point ELISA kit (Ansh Labs, USA).

[0088] Most subjects had their AMH levels measured on day 2 of their menstrual cycle, while a small number had their AMH levels measured at a time that was adjusted according to actual needs.

[0089] The data used to build the model is shown in Table 1 below.

[0090] Table 1. Clinical and biochemical data of subjects treated with GnRH antagonists

[0091] System Model Building In this embodiment, poor ovarian response and fewer than 5 oocytes (specifically 0, 1, 2, 3, or 4) are defined as outcome variables, and age and AMH level are the predictor variables. In this embodiment, the predictive model is constructed using data from 2017-2019, i.e., data from 16,327 subjects as the training set to initially build the model system of this application. Data from 2020-2021, i.e., data from 15,596 subjects, is used as the validation set to verify the effectiveness of the system model.

[0092] The probability of low ovarian response was modeled using a logistic regression framework for direct comparison with the applicant's previously published model (Model-0, as disclosed in CN2021104387761). Three candidate specifications for age and AMH were evaluated: Model-1 (Continuous): Age and AMH are included as untransformed continuous variables.

[0093] Model-2 (polynomial): Age is represented by a quadratic term, and AMH is represented by a cubic term, based on exploratory generalized additive modeling.

[0094] The model's discriminative power was quantified by the area under the receiver operating characteristic (ROC) curve (AUC), sensitivity, specificity, and F1 score. Calibration effectiveness was assessed by comparing predicted probabilities with observed POR rates using calibration plots. The incremental improvement compared to Model-0 was measured using the Net Reclassification Improvement (NRI) index.

[0095] The relationship between predicted low ovarian response rate (also defined as predicted DOR in this study) and age was then described. Given the expected S-shaped pattern, logistic growth curves were used to fit the age-DOR data.

[0096] All analyses were performed in JMP Pro 17.0 (SAS Institute) and R 4.4.1, and statistical significance was defined as two-sided p < 0.05.

[0097] Picture 1 The calibration performance of Model-0, Model-1, Model-2, and the ideal model was compared on the training (2017–2019) and testing (2020–2021) datasets. Model-2 performed best, with its curve closely aligned to the ideal model, especially in the low to medium probability range. While Model-1 also performed well, Model-0 showed significant differences compared to the other models. The following figure magnifies the 0–0.25 probability range, which is particularly important for representing the 82.3% female population in our dataset. In this magnified view, Model-2 again outperforms the other models, demonstrating superior calibration and making it the most reliable model for predicting outcomes within this most common probability range.

[0098] Based on these findings, we selected Model-2 as the final updated version. Model calibration is crucial in clinical practice. Proper calibration ensures that the predicted incidence of perimenopause accurately reflects the true incidence in the population, thus guaranteeing that each predicted value is close to the actual incidence. This accurate incidence estimate provides a solid foundation for subsequent predictions, such as the age at onset of perimenopause.

[0099] Simultaneously, the data from the aforementioned training group were used to fit the following calculation method to determine the probability of low ovarian response in the subjects ( p The formula for ) is: (Formula 1) As shown in Formula 1, p e represents the probability of low ovarian response, calculated to characterize the subject's current ovarian reserve function; Where a is any value selected from -3.830 to -2.929, b is any value selected from 0.042 to 0.069, c is any value selected from 0.0017 to 0.0050, d is any value selected from -1.490 to -1.327, f is any value selected from 0.250 to 0.373, and g is any value selected from 0.091 to 0.130.

[0100] Determining the ovarian aging curve and endocrine age In this study, we aimed to further optimize the ovarian aging curve by incorporating more data and exploring additional methods. First, we recognized that ovarian aging follows an S-shaped trajectory. Therefore, we applied a two-parameter logistic curve to describe the predicted ovarian rate of return (POR) at different ages (i.e., the POR calculated using the formula above). p The proportion trend of DOR. POR is based on a newly developed Model-2 definition, which applies a third transformation to AMH. Using an actual POR incidence rate of 0.15 as the threshold for defining POR, we plotted the proportion of DOR across different age groups. Picture 2 The logistic curve model implements r. 2 The improvement of 0.991, which is better than the original 0.978, proves that the overall fit is better.

[0101] Using a fitted S-shaped curve, we can map any individual's predicted ovarian reserve (POR) probability onto an ovarian aging curve to derive their "endocrine age." Specifically, we calculate the predicted POR probability by measuring AMH levels and recording the subject's chronological age. We then find the position corresponding to this probability on an S-shaped ovarian aging curve constructed from large-scale population data; the age at which the curve reaches this probability is defined as the ovarian endocrine age. If a young woman's ovarian reserve score corresponds to an endocrine age significantly higher than her chronological age, this may indicate accelerated ovarian decline and a possible earlier decline in fertility. Conversely, if her endocrine age is close to or lower than her chronological age, it indicates that her ovarian function is relatively well preserved. In this way, ovarian endocrine age reflects the true state of ovarian reserve, revealing the degree and rate of ovarian aging more accurately than chronological age.

[0102] Based on the probability of low ovarian response, endocrine age can be calculated by back-calculating from a two-parameter logistic curve using the following formula (Formula 2). Endocrine ): age Endocrine = (Formula 2) Where a1 and b1 are estimated values ​​of the inflection point and growth rate parameters obtained based on the logic curve, and are unitless parameters; where a1 is a fixed value of 37.553 and b1 is a fixed value of 0.283.

[0103] The standard error for calculating endocrine age ( se age This allows for the calculation of endocrine age. Endocrine The 95% confidence interval of ) se age = in, p The probability of low ovarian response for ovarian reserve function, where a2, b2, and c2 are unitless parameters; where a2 is a fixed value of 0.283; b2 is a fixed value of 0.012; and c2 is a fixed value of 0.168. The lower and upper limits of the 95% confidence interval for calculating endocrine age are as follows: age Endocrine_lower = age Endocrine -1.96 se age age Endocrine_upper = age Endocrine +1.96 se age .

[0104] Age of onset for predicting future reproductive milestones The "fixed-interval hypothesis" posits that an individual's ovarian aging trajectory follows the same curve as the population average—meaning the time intervals between different stages of ovarian decline are constant, only the starting points differ. In other words, although individuals may begin with different levels of ovarian reserve, the overall pattern of functional decline over time can be described by a single S-shaped curve. Evidence supporting this hypothesis primarily comes from cross-sectional observations. This hypothesis provides the theoretical basis and methodological support for predicting future changes in ovarian reserve from a single measurement.

[0105] Building upon this, the time intervals required for an individual to reach a specific ovarian reserve state (predicted probability of return) were derived. First, the subject's current ovarian reserve score was mapped onto an S-shaped curve to determine its position within the overall aging process. Then, using the fitted curve, the intervals required to drop from the current score to the predetermined probability of return (e.g., the specific probability of return corresponding to a specific ovarian reserve score) were calculated, and these intervals were converted into the age required to reach each milestone.

[0106] Specifically, Formula 3 can be used to calculate the number of years from the subject's current ovarian reserve to the point where a significant decline in ovarian reserve leads to a significant decline in fertility, i.e., the number of years when the probability of low ovarian response is 50%: Based on endocrine age, the number of years until ovarian reserve reaches 50 points (D1) is calculated, which is obtained from the logistic curve. p =0.5 times the difference between age and endocrine age; D1=age( )-age Endocrine =a1 - age Endocrine (Formula 3) p The probability of low ovarian response for ovarian reserve function; a1 and b1 are estimated values ​​of the inflection point and growth rate parameters obtained based on the logistic curve, and are unitless parameters; a1 is 37.553; b1 can be any value from 0.260 to 0.301, and is further preferred to be 0.283; age Endocrine Any value between the lower and upper limits of the 95% confidence interval in Formula 2 can be used to further optimize age. Endocrine This is the calculated value from Formula 2.

[0107] The age at which ovarian reserve reaches 50 points is calculated using the following formula. DOR ): age DOR = age chronological + D1 Among them, age chronological This represents the subject's current age, and D1 is the D1 value calculated using Formula 3.

[0108] Furthermore, Formula 4 can be used to calculate the number of years from the subject's current ovarian reserve to the onset of perimenopause, i.e., the number of years when the probability of low ovarian response is 95%: Based on endocrine age, the number of years to perimenopause (D2) is calculated, which is obtained from the logistic curve. p The difference between age and endocrine age at a value of 0.95: D2=age( )-age Endocrine =a1 - age Endocrine (Formula 4) pThe probability of low ovarian response for ovarian reserve function; a1 and b1 are estimates of the inflection point and growth rate parameters obtained based on the logistic curve, and are unitless parameters; a1 can be any value from 37.224 to 37.883, with a1 being more preferably 37.553; b1 can be any value from 0.260 to 0.301, with b1 being more preferably 0.283; age Endocrine Any value between the lower and upper limits of the 95% confidence interval in Formula 2 can be used, further aging. Endocrine The calculated value of Formula 2 is preferred.

[0109] The age at the onset of perimenopause can be calculated using the following formula: age 围绝经 = age chronological + D2 Among them, age chronological D represents the subject's current age, and D2 is the D2 value calculated using Formula 4.

[0110] Example 2: Selection of Blood Collection Time Point While most women have relatively stable AMH levels within a cycle, some exhibit significant intra-cycle fluctuations. To assess the impact of these variations on predictive performance, we compared AMH measurements on days 2, 6, and the hCG trigger day (approximately day 12) within a GnRH antagonist cycle and constructed three different models using AMH and age to determine the extent to which the magnitude of AMH decline significantly affects model accuracy.

[0111] First, we performed a normalization transformation on the non-normally distributed AMH values ​​( Picture 3 A). Then, using a linear mixed-effects model of the log-transformed data, we found that compared to AMH on day 2 of the cycle (AMH2), AMH on day 6 (AMH6) and day 12 (AMH2) were significantly lower. 12 The AMH levels of these two groups decreased by approximately 17.4% and 49.7%, respectively (p < 0.0001 for both). Picture 3 B) indicates that AMH levels peak during the early follicular phase and then decline significantly.

[0112] ROC curve analysis showed that the model including AMH and age on day 2 of the cycle achieved the highest AUC (0.868; 95% CI, 0.814–0.908). Although its AUC was not significantly different from the model including AMH and age on day 6 of the cycle (AUC = 0.860; 95% CI, 0.805–0.902), it was significantly better than the model including AMH and age on day 12 of the cycle (AUC = 0.652; 95% CI, 0.584–0.715). Picture 3 C). Furthermore, the confusion matrix indicators showed that the day 2 AMH and age model had the lowest misclassification rate (0.1099) and outperformed other models in terms of sensitivity, F1 score, and Matthews correlation coefficient (MCC), indicating that the model is more accurate and reliable in predicting POR in women with large fluctuations in AMH cycles.

[0113] Therefore, it is recommended that the model for predicting ovarian aging in the subjects of this application be based on the highest AMH concentration, because at this concentration, AMH best reflects the state of ovarian reserve, i.e., it is least affected by the negative regulation of E2.

[0114] Although the embodiments of this application have been described above, this application is not limited to the specific embodiments and application fields described above. The specific embodiments described above are merely illustrative and instructive, and not restrictive. Those skilled in the art can make many other forms based on the guidance of this specification and without departing from the scope of protection of the claims of this application, and these are all within the scope of protection of this application.

Claims

1. A system for predicting ovarian aging in a subject, comprising: The data acquisition module is used to acquire data on the age and anti-Müllerian hormone (AMH) levels of the subjects; The module for calculating ovarian low-response function is used to calculate the above data acquired in the data acquisition module, thereby calculating the probability (p) of ovarian low-response of the subject, and further converting the probability of ovarian low-response into a standardized ovarian reserve score. A module for calculating a subject's endocrine age, which uses the probability of a subject's low ovarian response to calculate the subject's endocrine age; and The module predicts the age at which changes in a subject's ovarian reserve occur, using the subject's endocrine age to predict the time or age at which new changes in ovarian reserve will occur. In the module for calculating ovarian hyporesponsiveness, the probability (p) of ovarian hyporesponsiveness is calculated by performing a quadratic transformation on the subject's age and a cubic transformation on the subject's anti-Müllerian hormone (AMH) level data. Calculate the probability of low ovarian response in the subject ( p The formula for () is as follows: Formula 1: (Formula 1) Where e represents the natural constant; Where a is any value selected from -3.830 to -2.929, b is any value selected from 0.042 to 0.069, c is any value selected from 0.0017 to 0.0050, d is any value selected from -1.490 to -1.327, f is any value selected from 0.250 to 0.373, and g is any value selected from 0.091 to 0.130; age represents the subject's age, and AMH represents the subject's anti-Müllerian hormone (AMH) level; In the module for calculating ovarian reserve function, the subject's age and anti-Müllerian hormone (AMH) level are logarithmic and substituted into Formula 1 for calculation.

2. The system according to claim 1, wherein, For women with regular menstrual cycles, the anti-Müllerian hormone (AMH) level refers to the concentration of anti-Müllerian hormone in the venous blood of the female subject on any given day. For women with irregular menstruation, the anti-Müllerian hormone (AMH) level refers to the highest concentration of anti-Müllerian hormone in venous blood during multiple consecutive venous blood draws in a female subject's menstrual cycle.

3. The system according to claim 1, wherein, Low ovarian response refers to a subject retrieving fewer than 5 eggs.

4. The system according to claim 1, wherein, In the module for calculating ovarian hyporesponsiveness, the probability (p) of ovarian hyporesponsiveness of a subject was calculated based on a logistic regression model using the subject's age and anti-Müllerian hormone (AMH) level.

5. The system according to claim 1, wherein, In the module for calculating low ovarian response function, the ovarian reserve score is calculated using the following conversion formula: Ovarian reserve score = (1 - probability of low ovarian response) × 100.

6. The system according to claim 1, wherein, In the module for calculating the subject's endocrine age, the subject's endocrine age is calculated using the probability of low ovarian response, based on any one of the following: a two-parameter logistic regression model, a two-parameter Probit model, a three-parameter Gompertz model, or a Weibul curve.

7. The system according to claim 6, wherein, The threshold is set to 0.15 in two-parameter logistic regression models, two-parameter Probit models, three-parameter Gompertz models, or Weibul curves.

8. The system according to claim 1, wherein, The calculated probability of poor ovarian response and standardized ovarian reserve score for the subjects included the following five categories: An ovarian low response rate of less than 5% is converted to an ovarian reserve score of greater than 95. An ovarian reserve score of 5% or higher and less than 25% is converted to an ovarian reserve score of 75 or higher and less than or equal to 95. An ovarian reserve score of 25% to 75% or less is equivalent to an ovarian reserve score of 25% to 50%. An ovarian reserve score of 50% to 50% with a low ovarian response rate is converted to an ovarian reserve score of 25 to 50. An ovarian reserve score of 75% to 25% is equivalent to an ovarian reserve score of 75% to 84%.

9. The system according to claim 5, wherein, In the module for calculating the subject's endocrine age, based on the probability of low ovarian response, the endocrine age (age) is calculated by back-calculating from a two-parameter logistic curve using Formula 2 as follows. Endocrine ): age Endocrine = (Formula 2) Where a1 and b1 are estimated values ​​of the inflection point and growth rate parameters obtained based on the logic curve, and are unitless parameters; Where a1 is a fixed value; b1 is a fixed value.

10. The system according to claim 9, wherein, In the module for calculating the subject's endocrine age, the standard error of the endocrine age is calculated ( se age This allows for the calculation of endocrine age. Endocrine The 95% confidence interval of ) se age = Where a2, b2, and c2 are unitless parameters; Where a2 is a fixed value of 0.283; b2 is a fixed value of 0.012; and c2 is a fixed value of 0.

168. The lower and upper limits of the 95% confidence interval for calculating endocrine age are as follows: age Endocrine_lower = age Endocrine -1.96 se age age Endocrine_upper = age Endocrine +1.96 se age 。 11. The system according to claim 1, wherein, In the module predicting the age at which a subject's ovarian reserve changes, the following formula (Formula 3) is used to calculate the number of years from the subject's current ovarian reserve status to the point where a significant decline in ovarian reserve leads to a significant decline in fertility, i.e., the number of years when the probability of low ovarian response is 50%: Based on endocrine age, the number of years until ovarian reserve reaches 50 points (D1) is calculated, which is obtained from the logistic curve. p =0.5 times the difference between age and endocrine age; D1=age( )-age Endocrine =a1 - age Endocrine (Official 3) a1 and b1 are estimated values ​​of the inflection point and growth rate parameters obtained based on the logic curve, and are unitless parameters; a1 is any value between 37.224 and 37.883; b1 is any value between 0.260 and 0.301; age Endocrine It is any value between the lower and upper limits of the 95% confidence interval in Formula 2.

12. The system according to claim 11, wherein, The age at which ovarian reserve reaches 50 points is calculated using the following formula. DOR ): age DOR = age chronological + D1 Among them, age chronological This represents the subject's current age, and D1 is the D1 value calculated using Formula 3.

13. The system according to claim 1, wherein, In the module predicting the age at which a subject's ovarian reserve changes, the following formula (Formula 4) is used to calculate the number of years from the subject's current ovarian reserve to the onset of perimenopause, i.e., the number of years when the probability of low ovarian response is 95%: Based on endocrine age, the number of years to perimenopause (D2) is calculated, which is obtained from the logistic curve. p The difference between age and endocrine age at a value of 0.95: D2=age( )-age Endocrine =a1 - age Endocrine (Official 4) a1 and b1 are estimated values ​​of the inflection point and growth rate parameters obtained based on the logic curve, and are unitless parameters; a1 is any value between 37.224 and 37.883; b1 is any value between 0.260 and 0.301; age Endocrine It is any value between the lower and upper limits of the 95% confidence interval in Formula 2.

14. The system according to claim 13, wherein, The age at the onset of perimenopause is calculated using the following formula (age). 围绝经 ): age 围绝经 = age chronological + D2 Among them, age chronological D represents the subject's current age, and D2 is the D2 value calculated using Formula 4.

Images