Hyperopic reserve prediction modeling method and computer system

Abstract

The application discloses a presbyopia reserve prediction modeling method, which comprises the following steps: S1: obtaining a first measurement model in overall data according to the overall data of a first sample pool through first processing; obtaining sample data sets from the overall data multiple times, and obtaining multiple second measurement models through second processing on each sample data set; the first measurement model comprises first structure parameters and a first error term; the second measurement models comprise multiple second structure parameters and second error terms; S2: constructing a structure parameter estimation model of the first measurement model and the second measurement models, obtaining an optimal unbiased estimator to determine the first measurement model of the overall data as a presbyopia reserve prediction model. The application also relates to a computer system for presbyopia reserve prediction. The presbyopia reserve prediction modeling method realized according to the application can improve the prediction accuracy of presbyopia reserve by using a computer system.

Description

Technical Field

[0001] This invention relates to the field of refractive power measurement technology, specifically to a method and computer system for predicting hyperopia reserve. Background Technology

[0002] In recent years, some researchers have proposed the concept of hyperopic reserve in children, which represents the sum of latent and manifest hyperopia. As children grow older, the hyperopic reserve gradually decreases until it reaches zero, which is the process of emmetropization. The level of the hyperopic reserve and the rate of its decrease determine the length of the emmetropization process. Therefore, understanding the hyperopic reserve in children is of great guiding significance for myopia prevention and control.

[0003] Because the amount of latent hyperopia varies among children, and the younger the child, the greater the proportion of latent hyperopia in their total hyperopia. For example, two four-year-old children may have a refractive error of +0.50D in the non-cycloplegic state, but after cycloplegia, one child may still have +0.5D, while the other may have +4.0D. Clearly, the former has less hyperopic reserve, and given a similar rate of hyperopia reduction, the latter has a higher risk of myopia and should increase outdoor activities and reduce close-range eye use to prevent the development and progression of myopia. Therefore, directly using the refraction measurement in the non-cycloplegic state cannot provide a sufficiently accurate reference for myopia prevention and control.

[0004] Although cycloplegic refraction can obtain a more accurate hyperopic reserve, it is not routinely performed in clinical practice on children who are still in a non-cycloplegic state and have low hyperopia or emmetropia. This is because their visual functions and vision are normal, and cycloplegia can cause side effects such as blurred near vision and photophobia. However, for these children, understanding their hyperopic reserve is more instructive for myopia prevention.

[0005] Some studies have attempted to predict true refractive error using refraction measurements taken in a non-cycloplegic state, but the predictive ability is very low. Other studies have focused on changes in axial length and, taking into account corneal curvature, proposed using axial ratio to predict the risk of myopia, but this does not directly reveal the patient's hyperopic reserve.

[0006] Therefore, there is an ongoing need in this field to develop a method and computer system for predicting farsighted reserves. Summary of the Invention

[0007] The purpose of this invention is to provide a method and computer system for predicting hyperopia reserve. This invention establishes the difference in refractive error of users before and after cycloplegia through a sample database, and mines the data features in the sample database through big data analysis to obtain the relationship between the refractive error values ​​before and after cycloplegia and various factors and elements. In particular, it models and predicts hyperopia reserve by using factors such as gender, height, axial length, corneal curvature, etc., to provide more accurate prediction values ​​for the hyperopia reserve of new users in a non-paralyzed state, and to provide scientific and convenient guidance for myopia prevention and control in clinical practice.

[0008] This invention discloses a method for predicting hyperopia reserves, characterized by the following steps:

[0009] S1: Based on the overall data of the first sample pool, the first econometric model is obtained from the overall data through the first processing;

[0010] Multiple sample datasets are obtained from the overall data, and multiple second econometric models are obtained by performing a second processing on each sample dataset;

[0011] The first measurement model includes a first structural parameter and a first error term;

[0012] The second econometric model includes multiple second structural parameters and a second error term;

[0013] S2: Construct the structural parameter estimation model of the first econometric model and the second econometric model, obtain the optimal unbiased estimator, and determine the first econometric model of the total data as the farsighted reserve prediction model.

[0014] Furthermore, the method further includes the following steps after step S2:

[0015] S3: Perform a test on the hyperopia reserve prediction model based on the test model.

[0016] Further, step S3 includes the following steps:

[0017] S31: Test the degree of fit based on the goodness-of-fit model and the total data;

[0018] S32: Based on the first significance test model, the linear relationship between the independent variable and the dependent variable in the hyperopia reserve prediction model based on the determined first structural parameter;

[0019] S33: The influence of the independent variable based on the determined first structural parameter on the dependent variable in the hyperopia reserve prediction model based on the second significance test model.

[0020] Furthermore, the first measurement model is:

[0021]

[0022] Where (x) p ,y p ) represents the data sampled from the first sample pool, where p is the sample number within the corresponding sample pool. The sample regression equation is obtained from the sampled data. The sample of the independent variable is used for sampling. Here, e represents the structural parameters of the sample corresponding to the independent variable, and e0 represents the obtained residual.

[0023] Furthermore, the second econometric model is:

[0024] Where X is the sample value matrix of the independent variable and Y is the sample value matrix of the dependent variable.

[0025] Furthermore, the hyperopia reserve prediction model is as follows:

[0026] CSE=a+b*Sex+c*Age+d*Height-e*AL+f*CR;

[0027] CSE is the farsightedness reserve value, Sex is the gender quantification value, Age is the age, Height is the user's height, AL is the axial length measurement value, and CR is the corneal curvature parameter.

[0028] This invention also discloses a computer system for predicting farsightedness reserve, characterized in that,

[0029] The computer system includes:

[0030] Memory used to store instructions and data, and

[0031] A processor coupled to the memory executes the memory instructions to perform the following operations:

[0032] Retrieve user data based on API;

[0033] The data includes: the user's age, height, axial length, gender, corneal radius value, and axial length;

[0034] Based on some or all of the data, perform analysis to obtain the user's farsightedness reserve value, and

[0035] The hyperopia reserve value is used to instruct the user on an eye control scheme.

[0036] Furthermore,

[0037] The analysis involves using a first model to calculate the hyperopia reserve value based on the data.

[0038] The first model is: CSE = a + b*Sex + c*Age + d*Height - e*AL + f*CR;

[0039] CSE is the farsightedness reserve value, Sex is the gender quantification value, Age is the age, Height is the user's height, AL is the axial length measurement value, and CR is the corneal curvature parameter.

[0040] Furthermore,

[0041] The analysis involves storing the data and the farsightedness reserve value into the corresponding memory sample library, and updating the first model.

[0042] The hyperopia reserve value is calculated based on the updated first model using the data.

[0043] Furthermore,

[0044] Using the hyperopia reserve value to instruct the user's eye control scheme includes one or more of the following:

[0045] Ophthalmic lens parameters, contact lens parameters, orthokeratology parameters, drug treatment plan, or outdoor activity plan.

[0046] Furthermore,

[0047] The analysis involves storing the data and the hyperopia reserve value into corresponding memory sample libraries.

[0048] Based on the univariate linear correlation analysis, the hyperopia reserve value determined by each of the data types is obtained, and one of them is selected for output.

[0049] Furthermore,

[0050] The acquisition of the first model is as follows: based on the memory sample library, user data types for the establishment of the first model are screened using univariate correlation analysis; and the relationship between the farsightedness reserve value and the screened user data types is obtained based on multivariate correlation analysis.

[0051] Furthermore,

[0052] The coefficient value is:

[0053] a=11.425, b=0.368, c=0.008, d=0.005, e=1.747, f=3.518.

[0054] This invention also discloses a method for predicting hyperopia reserve, characterized in that the method includes the following steps:

[0055] Obtain user data; the data includes: user's age, height, axial length, gender, corneal radius value, and axial length.

[0056] Based on the data, the user's farsightedness reserve value is obtained through analysis using a first model, and

[0057] The control scheme for indicating the user's farsightedness reserve is used to indicate the farsightedness reserve value;

[0058] The first model is: CSE = a + b*Sex + c*Age + d*Height - e*AL + f*CR;

[0059] The CSE is the farsightedness reserve value, the Sex is the gender quantification value, the Age is the age, the Height is the user's height, the AL is the axial length measurement value, and the CR is the corneal curvature parameter. In the first model, the coefficient parameters are: a = 11.425, b = 0.368, c = 0.008, d = 0.005, e = 1.747, f = 3.518.

[0060] Furthermore, the acquisition of the first model involves: selecting user data types for the establishment of the first model based on the memory sample library using univariate correlation analysis; and obtaining the relationship between the farsightedness reserve value and the selected user data types based on multivariate correlation analysis.

[0061] Compared with the prior art, the positive effects of the present invention are as follows:

[0062] (1) This invention first proposes a modeling method for user eye data to establish a calculation model for predicting hyperopia reserve value. It proposes a modeling method based on univariate regression and multivariate regression, and performs various detection and verification during the model building process to obtain a predictive model for hyperopia reserve value.

[0063] (2) This invention proposes a screening method for associated user data based on actual user measurements, and trains a model to construct a relationship between associated type parameters, proposing a prediction model for hyperopia reserve and providing data-driven guidance for myopia prevention and control; the hyperopia reserve prediction model implemented according to this invention is as follows:

[0064] CSE=11.425+0.368*Sex+0.008*Age+0.005*Height-1.747*AL+3.518*CR(R 2 =0.804)

[0065] The method and system for predicting hyperopia reserve of the present invention are particularly suitable for children aged 3-10 years, and especially for obtaining more accurate predicted values ​​of hyperopia reserve using a computer system, which facilitates the generation of eye prescriptions and the management of hyperopia reserve values ​​as the child grows and various parameters change. Attached Figure Description

[0066] Figure 1 This diagram illustrates a flow chart of a farsighted reserve prediction modeling method according to the present invention. Detailed Implementation

[0067] Unless otherwise defined, the technical or scientific terms used in this specification and claims shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0068] The technical solution of the present invention will now be clearly and completely described in conjunction with the accompanying drawings and embodiments thereof.

[0069] Furthermore, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0070] The hyperopic reserve referred to in this invention is the sum of latent and manifest hyperopia in children, and is generally obtained during refraction testing under cycloplegic conditions. Firstly, hyperopic reserve is a value that changes with age and development, making it difficult to measure. Secondly, even after obtaining a measurement of hyperopic reserve, the implementation of these plans, and the subsequent assessment of any improvement in hyperopic reserve, cannot be obtained through measurement.

[0071] The model construction in this invention includes constructing a first sample pool of data. The data in the first sample pool is constructed into a sample library containing refraction data information under cycloplegic conditions. The information in the sample library includes multi-dimensional user information data, including but not limited to:

[0072] The data obtained from a comprehensive and standardized examination process includes uncorrected visual acuity, retinoscopy or subjective refraction, corrected visual acuity, eye position, slit-lamp anterior segment examination, fundus examination, intraocular pressure, axial length and corneal curvature, height, refraction before cycloplegia, and refraction after cycloplegia.

[0073] In a specific implementation, the IOL Master 500 (Zeiss, Germany) is used to measure axial length and corneal curvature. The validity of the data is evaluated by the signal-to-noise ratio (SNR), where SNR ≥ 2 indicates the reliability of the data.

[0074] Based on the data from the first sample pool mentioned above, the following training parameters are further included in the construction of the execution model:

[0075] The spherical equivalent (SE) is equal to the spherical lens plus half of the cylindrical lens (spherical lens + 0.5 * cylindrical lens);

[0076] The axial length / radius of curvature (AL / CR) is equal to the ratio of the axial length of the eye to the radius of curvature of the cornea; where AL is the axial length of the eye and CR is the radius of curvature of the cornea.

[0077] Noncycloplegic spherical equivalent (NSE) is the refractive power of the ciliary muscle when it is not paralyzed.

[0078] Cycloplegic spherical equivalent (CSE) is the refractive power of the ciliary muscle when it is paralyzed.

[0079] The difference between noncycloplegic and cycloplegic spherical equivalent (DSE) is equal to CSE minus NSE.

[0080] The data in the first sample pool of this invention correspond to children aged 3-10 years. The selection of an equivalent spherical power of -3.0D or higher is based on the consideration that, due to accommodation, the refractive power of children with low myopia will decrease after cycloplegia. Hyperopic reserve is the sum of manifest and latent hyperopia, and must be obtained under cycloplegia. In this invention, CSE represents the degree of hyperopic reserve. In other embodiments, DSE can also be used to obtain the relationship between assessment parameters in hyperopic reserve for modeling.

[0081] like Figure 1 As shown in the figure, the present invention proposes a method for predicting hyperopia reserves, which includes the following steps:

[0082] S1: Based on the overall data of the first sample pool, the first econometric model is obtained from the overall data through the first processing;

[0083] The overall data of the first sample pool in this invention consists of various detection data surrounding the eyeball and the user's body, including gender, height, age, AL and CR.

[0084] In one embodiment of the present invention, the first processing involves using the attributes of the user's test data as independent variables and CSE or DSE as the hyperopic reserve prediction value as the dependent variable. The first measurement model is used to determine the relationship between the independent and dependent variables. Thus, using the sampled values ​​of the aforementioned dependent variables of a new user, the hyperopic reserve prediction value calculated by the algorithm can be obtained. This can generate an ophthalmic prescription for the user, including but not limited to ophthalmic lens parameters, contact lens parameters, corneal reshaping parameters, drug treatment plans, or outdoor activity plans. Furthermore, after the user has received treatment according to the prescription, based on the collected values ​​of the new independent variables, a new hyperopic reserve prediction value for the user can be obtained, thereby providing feedback to correct the first measurement model.

[0085] Specifically, the first econometric model established is as follows:

[0086] Y = Xβ + U, where U is the first error term and β is the first structural parameter. Due to the existence of the random error term U, Y (dependent variable) and X (independent variable) are not on the same straight line / plane. Since we assume EU = 0, the mean of the dependent variable and the independent variable are always on the same straight line. The straight line E(Y|X) = Xβ is the overall regression line (equation).

[0087] Multiple sample datasets are obtained from the overall data, and multiple second econometric models are obtained by performing a second processing on each sample dataset;

[0088] Specifically, the second econometric model established is as follows:

[0089] Estimated from sample data The relationship equation between the fitted values ​​of a portion of the sample observations and the independent variables was obtained. This is called the sample regression equation. The second econometric model mentioned above is obtained based on the sampled data, where e is the residual. Since there are multiple possibilities for the sampled data, the estimated value obtained from each sampling is... They will all be different, that is, the estimators of β. It is a random variable.

[0090] The establishment of the above-mentioned econometric model mainly involves training with sample data to construct the relationship between independent and dependent variables, thereby constructing an econometric process to determine which structural parameters the dependent variable is related to under various parameter conditions.

[0091] S2: Construct structural parameter estimation models for the first and second econometric models, obtain the optimal unbiased estimator, and use the first econometric model for overall data as the long-sighted reserve prediction model.

[0092] The structural parameter estimation models include, but are not limited to, univariate regression models and multivariate regression models. The steps for obtaining the optimal unbiased estimator using the structural parameter estimation model include:

[0093] (1) Obtain multiple second structure parameters using univariate regression model and multivariate regression model respectively;

[0094] The univariate regression model is:

[0095]

[0096] Where (x) p ,y p The data is sampled from the first sample pool, where p is the sample number within the first sample pool. In one implementation, the total sample population is m, and the structural parameter dimension of each number may be t. In the univariate regression model, the main goal is to realize the relationship between the sample regression equation and each structural parameter, thereby obtaining the institutional parameters related to the final hyperopia reserve value. This allows for the selection of a subset of structural parameters from a large number of types, corresponding to the relationship between CSE or DSE and the corresponding independent variable. The sample regression equation is obtained from the sampled data. The sample of the independent variable is used for sampling. Here are the structural parameters of the sample corresponding to the independent variable, and e0 is the corresponding residual obtained;

[0097] (2) Based on the structural parameters selected from the univariate regression model and the corresponding sample values, the total number of sample values ​​is m, and the dimension of the structural parameters of the independent variables is l, where l≤t. Further, multiple second structural parameters are obtained using the bivariate regression model. At this time, the bivariate regression model includes multiple samples and corresponding equations. For example, if 2 structural parameters are selected, or 3 structural parameters are selected, multiple second structural parameters can be obtained using the following multivariate regression model. This number is n, and n sample equations can be obtained accordingly:

[0098] The multiple regression model is as follows:

[0099] And obtain the corresponding residual term e, where the independent variable X is a matrix with multiple sample parameters.

[0100] S3: For the multiple population regression equations and multiple sample regression equations obtained in step S2 above, filter them to obtain the optimal unbiased estimator. The filtering conditions are as follows:

[0101] ① The expected value of the random error term is zero, E(U). i =0; that is, to use all the residuals obtained from the regression equation of the acquired sample as the expected value for calculation;

[0102] ② The random error term has homoscedasticity Var(u) i )=σ 2 i = 1, 2, ..., n;

[0103] ③ The random error terms are uncorrelated with each other (Cov(u)). i ,u j )=0 i≠j;i,j=1,2,…,n;

[0104] ④ Explain the variables X1, X2, ..., X k As a deterministic variable, it is uncorrelated with the random error term.

[0105] Cov(X ij ,u j )=0 i=1,2,…,kj=1,2,…,n;

[0106] ⑤ Explain the variables X1, X2, ..., X k There is no exact (complete) linear relationship between them, meaning the sample observation matrix X of the explanatory variables is a full-rank matrix: rank(X) = k+1 < n

[0107] ⑥ The random error term follows a normal distribution, that is:

[0108] As a preferred embodiment of this invention, the present invention further includes step S4:

[0109] S41: Test the goodness of fit between the goodness-of-fit model and the total data to test the goodness of fit of the regression equation to the sample points;

[0110] In one embodiment of the present invention, the goodness-of-fit model is as follows:

[0111]

[0112]

[0113] in,

[0114]

[0115] Among them, Y i For true sample values, The fitted dependent variable function value of the sample regression function. The value of the dependent variable in the overall regression function.

[0116] S42: Based on the first significance test model, the linear relationship between the independent and dependent variables in the hyperopia reserve prediction model based on the determined first structural parameter is established:

[0117] The first hypothesis under the first significance test model is given as follows:

[0118] Constructing statistics:

[0119] Given a significance level α, determine the rejection region F > F. α (k,nk-1);

[0120] Calculate the statistical test result and determine whether to reject the first hypothesis. In this way, test whether a significant linear relationship exists between the dependent and independent variables in the model. (Further details can be added if possible.)

[0121] S43: The influence of the independent variable based on the determined first structural parameter on the dependent variable in the hyperopia reserve prediction model is determined according to the second significance test model.

[0122] The second significance test model is: multiple regression: Where C i+1,i+1 For (X′X) -1 The element located in the (i+1)th row and (i+1)th column;

[0123] In one preferred embodiment of the present invention, during the univariate regression model calculation stage in step S1, the aforementioned second significance test model is used to verify the univariate regression model in the screening structural parameters, including the following second significance test model:

[0124] One-yuan return:

[0125] The basic steps of variable significance testing are as follows:

[0126] The second hypothesis is proposed: H0:β i =0 H1:β i ≠0;

[0127] Constructing statistics:

[0128] Given a significance level α, determine the rejection region |t|>t. α / 2 (nk-1);

[0129] Calculate the statistical value and determine whether to reject the null hypothesis, and test whether the independent variables in the model have a significant effect on the dependent variable.

[0130] Based on the above modeling method, this invention further proposes a method for predicting hyperopia reserve, which includes the following steps:

[0131] Obtain user data; the above data preferably includes: user's age, height, axial length, gender, corneal radius value, and axial length.

[0132] Based on the above data, the user's farsightedness reserve value is obtained through analysis using the first model;

[0133] The first model is: CSE = a + b*Sex + c*Age + d*Height - e*AL + f*CR;

[0134] CSE is the farsightedness reserve value, Sex is the gender quantification value, Age is the age, Height is the user's height, AL is the axial length measurement value, CR is the corneal curvature parameter, and gender is 1 for males and 0 for females; age is calculated in months; and height is calculated in centimeters.

[0135] That is, based on the modeling results, the first structural parameters are constructed as follows: (age, height, axial length, gender, corneal radius value, axial length).

[0136] Furthermore, the coefficients of the structural parameters are as follows:

[0137] CSE=11.425+0.368*Sex+0.008*Age+0.005*Height-1.747*AL+3.518*CR.

[0138] According to one specific embodiment of the present invention, the first model can be adjusted to CSE = a + b*Sex + c*Age + d*Height + g*AL / CR. Here, the coefficient g describes the relative relationship between AL and CR.

[0139] As age increases, the axial length of the eye gradually increases, the corneal curvature flattens, and the AL / CR ratio also gradually increases. Women, compared to men, have shorter axial lengths (AL) and steeper corneal curvatures (CR). In model construction and prediction, the AL / CR ratio showed a higher correlation with CSE than AL alone (AL:r = -0.7643, AL / CR:r = -0.8623). The AL / CR ratio provides a comprehensive factor of axial length and corneal refractive power, and compared to AL alone, AL / CR serves as a better predictor of myopia development. Furthermore, when the AL / CR ratio is greater than 3, the probability of developing myopia increases significantly.

[0140] According to another embodiment of the present invention, a method for acquiring farsightedness reserves is disclosed:

[0141] Obtain the first test dataset, which includes multiple attribute data parameters, including but not limited to age, gender, height, axial length, NSE, and CSE.

[0142] The following analysis steps were performed on multiple data points in the first test dataset to obtain a model for predicting hyperopia reserve:

[0143] The correlation between DSE and age and CSE was analyzed using univariate linear regression; or the correlation between CSE and AL and AL / CR was analyzed using univariate linear regression.

[0144] A prediction model for CSE was established using multiple linear regression analysis;

[0145] The second significance test model (t-test) was used to compare the differences in refractive power before and after cycloplegia.

[0146] As a specific embodiment of the present invention, the correlation between the refractive power of the left and right eyes before cycloplegia is (r = 0.871), and the correlation between the refractive power of the left and right eyes after cycloplegia is (r = 0.926). The data of the left and right eyes are highly correlated, and it is preferred to analyze the data of the right eye. The model can be constructed using the data of both eyes.

[0147] According to the first model of the present invention, the predicted values ​​of the actual values ​​are within ±0.5D of the predicted values ​​in 51.53% of cases, and the predicted values ​​are within ±1.0D of 86.26%.

[0148] According to a specific embodiment of the present invention, DSE does not decrease with increasing age; DSE has a low correlation with age (r = -0.4218) but a strong correlation with CSE.

[0149] Because hyperopic subjects require more accommodation when focusing on near targets during measurement, the difference in refractive error before and after cycloplegia is greater. Furthermore, when CSE ≥ -0.50D, the correlation between DSE and CSE is strong (r = 0.7889), while when CSE < -0.50D, the correlation is weak (r = 0.1586).

[0150] In clinical practice, children with emmetropia and low hyperopia have normal visual function, but due to cycloplegia, they may experience side effects such as photophobia and blurred vision. By establishing a predictive model using a sample database, a computer system can be used to obtain hyperopia reserve prediction values ​​in order to obtain eye control strategies.

[0151] The present invention further discloses a computer system for predicting farsightedness reserve, the computer system comprising:

[0152] Memory used to store instructions and data, and

[0153] A processor coupled to memory executes store instructions to perform the following operations:

[0154] Retrieve user data based on API;

[0155] The data includes: the user's age, height, axial length, gender, corneal radius, and axial length;

[0156] Based on partial or complete data analysis, the user's farsightedness reserve value is obtained, and

[0157] The farsightedness reserve value is used to indicate the user's eye control plan.

[0158] According to the prediction method and prediction system of the present invention, the level of hyperopic reserve is related to the refractive power after cycloplegia and has a higher correlation with AL / CR. When the refractive power after cycloplegia cannot be obtained, the prediction model can be used to predict hyperopic reserve, thereby providing certain guidance for clinical practice.

[0159] The above description of the embodiments is intended to enable those skilled in the art to understand and apply the present invention. It will be apparent to those skilled in the art that various modifications can be easily made to these embodiments, and the general principles described herein can be applied to other embodiments without creative effort. Therefore, the present invention is not limited to the embodiments described herein, and any improvements and modifications made by those skilled in the art based on the disclosure of the present invention without departing from the scope and spirit of the invention are within the scope of the present invention.

Claims

1. A method for predicting and modeling farsighted reserves, characterized in that, The method includes the following steps: S1: Based on the overall data of the first sample pool, the first measurement model in the overall data is obtained through the first processing; the overall data of the first sample pool includes the user's gender, age, height, axial length, and corneal curvature parameters, and the measurement data of the axial length and corneal curvature parameters must meet the signal-to-noise ratio SNR≥2 to ensure data validity; Multiple sample datasets are obtained from the overall data, and multiple second econometric models are obtained by performing a second processing on each sample dataset; The first measurement model includes a first structural parameter and a first error term. The first measurement model is as follows: ； in,( ) represents the data sampled from the first sample pool, where The sample number is the sample number in the first sample pool corresponding to the sampling, where The sample regression equation is obtained from the sampled data. The sample of the independent variable is used for sampling. For the structural parameters of the samples corresponding to the independent variables, This corresponds to the obtained residual; The second econometric model includes multiple second structural parameters and a second error term. The second econometric model is as follows: ； Where X is the sample value matrix of the independent variable and Y is the sample value matrix of the dependent variable; S2: Construct structural parameter estimation models for the first and second econometric models, obtain the optimal unbiased estimator, and determine the first econometric model for the total data as the hypersightedness reserve prediction model; the hypersightedness reserve prediction model is: CSE=a+b Sex+c Age+d Height-e AL+f CR; Alternatively, the following options may be available: CSE=a+b Sex+c Age+d Height+g AL / CR; Wherein, CSE is the hyperopia reserve value, Sex is the gender quantification value, Age is the age, Height is the user's height, AL is the axial length measurement value, CR is the corneal curvature parameter, AL / CR is the ratio of axial length to corneal curvature radius, and the coefficients satisfy: a=11.425, b=0.368, c=0.008, d=0.005, e=1.747, f=3.518, and g is the adjustment coefficient corresponding to AL / CR.

2. The hyperopia reserve prediction modeling method as described in claim 1, characterized in that, The method further includes the following steps after step S2: S3: Perform a test on the hyperopia reserve prediction model based on the test model.

3. The hyperopia reserve prediction modeling method as described in claim 2, characterized in that, Step S3 further includes the following steps: S31: Test the degree of fit based on the goodness-of-fit model and the total data; S32: The linear relationship between the independent variable and the dependent variable based on the determined first structural parameter in the hyperopia reserve prediction model is tested according to the first significance test model; S33: The influence of the independent variable based on the determined first structural parameter on the dependent variable in the hyperopia reserve prediction model is tested according to the second significance test model.

4. A computer system for predicting farsightedness reserve, characterized in that, The computer system includes: Memory used to store instructions and data, and A processor coupled to the memory executes the memory instructions to perform the following operations: Retrieves user data based on an API; The data includes: the user's age, height, axial length, gender, corneal radius, and corneal curvature parameters, and the measurement data of the axial length and corneal curvature parameters satisfy SNR≥2; Based on part or all of the data, an analysis is performed to obtain the user's hyperopia reserve value. The analysis is: using a first model to calculate the hyperopia reserve value through the data. The first model is: CSE = a + b Sex+c Age+d Height-e AL+f CR; Alternatively, the following options may be available: CSE=a+b Sex+c Age+d Height+g AL / CR; Wherein, CSE is the hyperopia reserve value, Sex is the gender quantification value, Age is the age, Height is the user's height, AL is the axial length measurement value, CR is the corneal curvature parameter, AL / CR is the ratio of axial length to corneal curvature radius, and the coefficients satisfy: a=11.425, b=0.368, c=0.008, d=0.005, e=1.747, f=3.518, and g is the adjustment coefficient corresponding to AL / CR; And using the hyperopia reserve value to instruct the user's eye control scheme.

5. The computer system for predicting farsightedness reserve as described in claim 4, characterized in that, The analysis involves storing the data and the farsightedness reserve value into the corresponding memory sample library, and updating the first model. The hyperopia reserve value is calculated based on the updated first model using the data.

6. The computer system for predicting farsightedness reserve as described in claim 4, characterized in that, Using the hyperopia reserve value to instruct the user's eye control scheme includes one or more of the following: Ophthalmic lens parameters, contact lens parameters, orthokeratology parameters, drug treatment plan, or outdoor activity plan.

7. The computer system for predicting farsightedness reserve as described in claim 4, characterized in that, The analysis involves storing the data and the hyperopia reserve value into corresponding memory sample libraries. Obtain the hyperopia reserve value determined by each of the data types based on univariate linear correlation analysis, and select one of them for output.

8. The computer system for predicting farsightedness reserve as described in claim 4, characterized in that, The first model is obtained by: screening user data types for the establishment of the first model based on univariate correlation analysis according to the memory sample library; and obtaining the relationship between the farsightedness reserve value and the screened user data types based on multivariate correlation analysis.

9. A method for predicting hyperopia reserve, characterized in that, The method includes the following steps: Acquire user data; the data includes: user's age, height, axial length, gender, corneal radius value, and corneal curvature parameter, and the measurement data of the axial length and corneal curvature parameter satisfy SNR≥2; Based on the data, the user's farsightedness reserve value is obtained through analysis using the first model, and The control scheme for indicating the user's farsightedness reserve is used to indicate the farsightedness reserve value; The first model is: CSE = a + b Sex+c Age+d Height-e AL+f CR; Alternatively, the following options may be available: CSE=a+b Sex+c Age+d Height+g AL / CR; Wherein, CSE is the hyperopia reserve value, Sex is the gender quantification value, Age is the age, Height is the user's height, AL is the axial length measurement value, CR is the corneal curvature parameter, AL / CR is the ratio of axial length to corneal curvature radius, and the coefficients satisfy: a=11.425, b=0.368, c=0.008, d=0.005, e=1.747, f=3.518, and g is the adjustment coefficient corresponding to AL / CR.

10. The method for predicting hyperopia reserve according to claim 9, characterized in that, The first model is obtained by: screening user data types for the establishment of the first model based on univariate correlation analysis according to the memory sample library; and obtaining the relationship between the farsightedness reserve value and the screened user data types based on multivariate correlation analysis.

Images