A myopia risk prediction method and system based on an eye biological parameter joint model

Abstract

This invention relates to the field of medical software technology and discloses a method and system for predicting myopia risk based on a joint model of ocular bioparameters. The method includes: acquiring at least one measurement of ocular bioparameters of a target individual, wherein the ocular bioparameters include axial length and corneal curvature data used to calculate the axial ratio; processing the ocular bioparameter measurement data of the target individual using a pre-trained joint myopia risk prediction model to obtain myopia risk information for the target individual within a predetermined future time period; and outputting the myopia risk information for the target individual within the predetermined future time period. This invention shares random effects parameters between a longitudinal sub-model and a survival sub-model, enabling the calculation process of myopia risk to reflect the individual differences contained in the trajectory of axial ratio changes of the target individual.

Description

Technical Field

[0001] This invention relates to the interdisciplinary field of medical informatics, ophthalmic public health, and artificial intelligence, and in particular to a method and system for predicting myopia risk based on a joint model of ocular biological parameters. Background Technology

[0002] Myopia has become a major global public health problem. Once myopia develops, it is irreversible, and high myopia can lead to blindness. Currently, clinical diagnosis relies on uncorrected visual acuity and computer-assisted refraction at a single point in time, lacking the ability to predict future risks, resulting in delayed interventions. Therefore, establishing an early warning system based on measurable physiological indicators to identify high-risk individuals before the onset of myopia is a key breakthrough in prevention and control.

[0003] Joint models that simultaneously process longitudinal biomarkers (such as tumor markers and CD4 cell counts) and time-event outcomes (such as death and disease progression), linking the two sub-models through shared random effects, have been shown to outperform traditional methods. However, in the field of ophthalmology, especially in the specific scenario of myopia occurrence among primary and secondary school students, there are no reports of joint models built on large-scale longitudinal cohorts and commercialized as mobile early warning systems. Summary of the Invention

[0004] The main objective of this invention is to address the technical problem that there is currently no mobile early warning system based on large-scale vertical queue construction and productization in the existing technology. A myopia risk prediction method based on a joint model of eye biological parameters includes the following steps:

[0005] Acquire at least one measurement of ocular bioparameters of the target individual, the ocular bioparameters including axial length and corneal curvature data for calculating the axial ratio;

[0006] Based on the eye biometric measurement data of the target individual, the data is processed by a pre-trained joint prediction model for myopia risk to obtain the risk information of myopia occurrence of the target individual in the future within a preset time period.

[0007] The joint prediction model for myopia risk is constructed through the following steps:

[0008] Based on at least one measurement of ocular bioparameters of the target individual;

[0009] The joint model simultaneously performs longitudinal sub-model processing and survival sub-model processing. The longitudinal sub-model is used to characterize the trajectory of the change in the axial ratio of the target individual and the historical group over time, and the survival sub-model is used to characterize the risk of myopia. The longitudinal sub-model and the survival sub-model are associated through shared random effects parameters.

[0010] The longitudinal sub-model is a mixed-effects longitudinal model, which uses the axial ratio as the longitudinal response variable and expresses the axial ratio as a function of time. The mixed-effects longitudinal model includes at least fixed-effects parameters to describe the trend of population-level changes and random-effects parameters to describe the differences between the target individual and the historical population.

[0011] The random effects parameters include at least individual random intercept parameters and individual random slope parameters. The individual random intercept parameters are used to characterize the deviation of the target individual from the initial level of the axial ratio, and the individual random slope parameters are used to characterize the deviation of the target individual's axial ratio from the rate of change over time relative to the population mean rate of change.

[0012] The survival sub-model is a time-event analysis model, in which the time of the first occurrence of myopia is used as the survival time variable, and whether or not myopia occurs is used as the event state variable.

[0013] The risk function of the survival sub-model incorporates at least one covariate, which includes an axial ratio obtained from at least one ocular bioparameter measurement of the target individual or a derived parameter calculated from the axial ratio.

[0014] The shared random effects parameters, as latent variables, participate in the parameter estimation process of both the longitudinal sub-model and the survival sub-model, enabling the individual difference information in the longitudinal sub-model used to describe the trajectory of axial ratio change over time to be synchronously reflected in the risk function in the survival sub-model used to describe the risk of myopia.

[0015] The method further includes a data preprocessing step, which includes data integrity verification, outlier identification, and data format standardization.

[0016] The outlier identification is used to determine whether the axial length or corneal curvature data exceeds the preset reasonable range. Data identified as outliers will not be included in the calculation of the subsequent myopia risk joint prediction model.

[0017] The data format standardization process is used to convert the eye biometric measurement data of the target individual into a data structure format consistent with the training phase of the joint myopia risk prediction model.

[0018] In the step of obtaining at least one eye biometric measurement data of the target individual, when the target individual has only one eye biometric measurement data, the myopia risk joint prediction model estimates the trajectory of the target individual's axial ratio change over time based on the longitudinal distribution characteristics of the eye biometrics of the historical group and the eye biometric measurement data of the target individual, through the posterior distribution of the preset random effects parameter in the longitudinal sub-model.

[0019] A second aspect of the present invention provides a myopia risk prediction system based on a joint model of ocular bioparameters, comprising:

[0020] The data acquisition module is used to acquire at least one measurement of ocular bioparameters of the target individual, the ocular bioparameters including axial length and corneal curvature data for calculating the axial ratio;

[0021] The model processing module is used to process the eye bioparameter measurement data of the target individual by calling a pre-trained joint prediction model for myopia risk, so as to obtain the myopia risk information of the target individual in the future within a preset time period; wherein, the joint prediction model for myopia risk is constructed in the following manner:

[0022] Based on at least one ocular biometric measurement data of the target individual; longitudinal sub-model processing and survival sub-model processing are performed simultaneously through a joint model, wherein the longitudinal sub-model is used to characterize the trajectory of the change in axial ratio over time in the data of the target individual and historical groups, and the survival sub-model is used to characterize the risk of myopia, and the longitudinal sub-model and the survival sub-model are associated through shared random effects parameters;

[0023] The result output module is used to output the risk information of myopia occurrence for the target individual within the preset future time period.

[0024] A third aspect of the present invention provides an electronic device, comprising: a memory and at least one processor, wherein the memory stores instructions, and the memory and the at least one processor are interconnected via a circuit; the at least one processor invokes the instructions in the memory to cause the electronic device to perform the myopia risk prediction method based on the joint model of eye bioparameters described above.

[0025] A fourth aspect of the present invention provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the aforementioned method for predicting myopia risk based on a joint model of eye bioparameters.

[0026] The present invention has the following beneficial effects:

[0027] This invention introduces both a longitudinal sub-model and a survival sub-model within the same joint model framework, and establishes a correlation between them using shared random effects parameters. This allows longitudinal information on the change of axial ratio over time to participate in modeling synchronously with the time-event outcome of myopia occurrence, avoiding the information fragmentation problem caused by the independence of longitudinal analysis and risk analysis in existing technologies.

[0028] This invention relies solely on at least one measurement of the target individual's eye bioparameters to combine historical population data with longitudinal and survival sub-models for joint processing. This allows for the prediction of myopia risk in a target individual within a predetermined timeframe, even with limited follow-up or incomplete screening data, thus broadening the applicable scenarios for myopia risk prediction technology.

[0029] This invention enables the calculation of myopia risk to reflect the individual differences contained in the trajectory of the target individual's axial ratio change by sharing random effects parameters between the longitudinal sub-model and the survival sub-model. This avoids the parameter estimation bias that may occur when the longitudinal prediction results are introduced as exogenous variables into the survival model, and improves the statistical consistency of the risk prediction results. Attached Figure Description

[0030] Figure 1 This is a flowchart of the myopia risk prediction method of the present invention.

[0031] Figure 2 This is a screenshot of the interface of the present invention applied to an APP.

[0032] Figure 3 This is a schematic diagram of the myopia risk prediction model of the present invention. Detailed Implementation

[0033] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” or “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0034] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1-3 The first embodiment of the myopia risk prediction method based on a joint model of eye bioparameters in this invention includes:

[0035] Acquire at least one measurement of ocular bioparameters of the target individual, the ocular bioparameters including axial length and corneal curvature data for calculating the axial ratio;

[0036] Based on the eye biometric measurement data of the target individual, the data is processed by a pre-trained joint prediction model for myopia risk to obtain the risk information of myopia occurrence of the target individual in the future within a preset time period.

[0037] The joint prediction model for myopia risk is constructed through the following steps:

[0038] Based on at least one measurement of ocular bioparameters of the target individual;

[0039] The joint model simultaneously performs longitudinal sub-model processing and survival sub-model processing. The longitudinal sub-model is used to represent the trajectory of the change in the axis ratio over time for the target individual and the historical group. The survival sub-model is used to represent the risk of myopia. The longitudinal sub-model and the survival sub-model are associated through shared random effects parameters.

[0040] As one implementation scheme, the longitudinal sub-model is a mixed-effects longitudinal model, which uses the axial ratio as the longitudinal response variable and expresses the axial ratio as a function of time; wherein, the mixed-effects longitudinal model includes at least fixed-effects parameters for describing the trend of change at the population level, and random-effects parameters for describing the difference between the target individual and the historical population.

[0041] As one implementation scheme, the random effects parameters include at least an individual random intercept parameter and an individual random slope parameter. The individual random intercept parameter is used to characterize the deviation of the target individual from the initial level of the axial ratio, and the individual random slope parameter is used to characterize the deviation of the target individual's axial ratio from the rate of change over time relative to the population mean rate of change.

[0042] As one implementation scheme, the survival sub-model is a time-event analysis model, wherein the survival sub-model uses the time of the first occurrence of myopia as the survival time variable and whether or not myopia occurs as the event state variable;

[0043] The risk function of the survival sub-model incorporates at least one covariate, which includes an axial ratio obtained from at least one ocular bioparameter measurement of the target individual or a derived parameter calculated from the axial ratio.

[0044] As one implementation scheme, the shared random effects parameters are used as latent variables to participate in the parameter estimation process of both the longitudinal sub-model and the survival sub-model. This allows the individual difference information in the longitudinal sub-model used to describe the trajectory of the axial ratio change over time to be synchronously reflected in the risk function in the survival sub-model used to describe the risk of myopia.

[0045] As one implementation scheme, the interaction between school type and axis ratio is introduced into the survival sub-model, making the prediction of the myopia risk function more accurate.

[0046] As one implementation scheme, it also includes a data preprocessing step, which includes data integrity verification, outlier identification, and data format standardization.

[0047] The outlier identification is used to determine whether the axial length or corneal curvature data exceeds the preset reasonable range. Data identified as outliers will not be included in the calculation of the subsequent myopia risk joint prediction model.

[0048] The data format standardization process is used to convert the eye biometric measurement data of the target individual into a data structure format consistent with the training phase of the joint myopia risk prediction model.

[0049] As one implementation scheme, in the step of obtaining at least one eye biometric measurement data of the target individual, when the target individual has only one eye biometric measurement data, the myopia risk joint prediction model estimates the trajectory of the target individual's axial ratio change over time based on the longitudinal distribution characteristics of the eye biometrics of the historical group and the eye biometric measurement data of the target individual, through the posterior distribution of the preset random effects parameter in the longitudinal sub-model.

[0050] Specifically, the model construction of this invention is carried out using the following steps:

[0051] Training and evaluation of a joint model of myopia development. Based on 8 years of continuous screening and follow-up data of 5,000 primary and secondary school students, a joint model of mixed effects (LME) and Cox proportional hazards model was constructed to achieve simultaneous modeling and parameter sharing of longitudinal process and survival outcome.

[0052] The system strictly complies with the "Personal Information Protection Law of the People's Republic of China" and the "Regulations on the Protection of Children's Personal Information Online." All data transmission uses TLS 1.3 encryption, and sensitive student information is anonymized. Only parents can access their own children's data after real-name authentication. The model calculation process is completed in a Trusted Execution Environment (TEE). The system implements strict access control policies, and only authorized personnel (such as parents or designated doctors) can access the corresponding refractive error records.

[0053] Step 1: Data Source and Queue Construction

[0054] Data source: Annual refractive error screening database of primary and secondary school students in a district of Shanghai (2017-2024), covering all grades from kindergarten to high school.

[0055] Sample size: Approximately 5,000 people, with an average of 4.2 follow-ups per person, and a total of nearly 30,000 screening records.

[0056] Inclusion criteria: no myopia at baseline (equivalent spherical power SE>-0.5D) and at least two follow-up data.

[0057] Outcome definition: The age or time point at which first myopia occurs (uncorrected visual acuity <5.0 and SE <-0.5D), enabling the construction of "time-event" data in survival analysis.

[0058] Step 2: Key Variable System

[0059] The dependent variable in the longitudinal sub-model is the axial ratio (AL / CR, continuous type), which is a robust predictor of refractive status because it can eliminate individual differences in corneal curvature better than AL alone.

[0060] Longitudinal sub-model fixed effects: age, sex, BMI, baseline spherical lens, cylindrical lens, astigmatism axis, corneal curvature K1 / K2, school type (key / ordinary), place of origin (Shanghai, non-Shanghai).

[0061] Longitudinal submodel random effects: individual random intercept and random slope (allowing each student to have different baseline AL / CR values ​​and growth rates).

[0062] Survival sub-model covariates: age, sex, school type, baseline SE, baseline AL / CR, and parental myopia history.

[0063] Interaction effect: The interaction between school type and longitudinal processes of AL / CR to capture differences in refractive developmental trajectories under different educational pressures.

[0064] Step 3: Mathematical Expression of the Joint Model

[0065] Longitudinal model:

[0066]

[0067] in,

[0068] in, For individuals In time AL / CR predicted values;

[0069] , The average effect strength of each covariate on AL / CR;

[0070] , For individuals In time eigenvectors;

[0071] For individuals Deviation of baseline value from population mean;

[0072] For individuals The deviation of the rate of change from the population mean;

[0073] , where is the follow-up period (in years / months);

[0074] Survival model:

[0075]

[0076] Among them, benchmark risk This represents the shared time risk of all individuals;

[0077] Baseline covariates Baseline hazard ratios representing baseline covariates such as gender, place of origin, and school type;

[0078] Longitudinal forecast This indicates the dynamic correlation between axial ratio (AL / CR) and myopia risk.

[0079] Interactive items This indicates the moderating effect of school type on the axis ratio effect.

[0080] Step 4: Model Fitting and Evaluation

[0081] Use the R language's JM or JMbayes packages for maximum likelihood estimation and the EM algorithm.

[0082] Internal validation: 10-fold cross-validation to assess discrimination and calibration.

[0083] The interaction effect of the joint model was significant (P=0.003) and the Concordance index was 0.851 (0.828-0.874), which was significantly higher than that of the time-dependent Cox model (C-index = 0.813) and the traditional Cox model (C-index = 0.734), indicating that the joint modeling has better discriminative ability in predicting the risk of myopia progression in individuals. The calibration curve showed that the predicted probability was highly consistent with the actual observed event frequency, and the Hosmer-Lemeshow test showed P = 0.472, indicating that the model has good calibration.

[0084] Compared to the single Cox model, it improved by approximately 3.8 percentage points, and the predictive sensitivity was >75% in 1, 2, and 3 years before the onset of myopia. Predictive sensitivity = (number of individuals correctly predicted as high-risk for progression) / (total number of individuals who actually experienced myopia progression). See Tables 1 and 2 for details (comparison of C-index for each model (Delong test)).

[0085] Table 1

[0086]

[0087] Table 2

[0088]

[0089] The above describes the myopia risk prediction method based on a joint model of ocular bioparameters in the embodiments of the present invention. The following describes the myopia risk prediction system based on a joint model of ocular bioparameters in the embodiments of the present invention:

[0090] The data acquisition module is used to acquire at least one measurement of ocular bioparameters of the target individual, the ocular bioparameters including axial length and corneal curvature data for calculating the axial ratio;

[0091] The model processing module is used to process the eye bioparameter measurement data of the target individual by calling a pre-trained joint prediction model for myopia risk, so as to obtain the myopia risk information of the target individual in the future within a preset time period; wherein, the joint prediction model for myopia risk is constructed in the following manner:

[0092] Based on at least one ocular biometric measurement data of the target individual; longitudinal sub-model processing and survival sub-model processing are performed simultaneously through a joint model, wherein the longitudinal sub-model is used to characterize the trajectory of the change in axial ratio over time in the data of the target individual and historical groups, and the survival sub-model is used to characterize the risk of myopia, and the longitudinal sub-model and the survival sub-model are associated through shared random effects parameters;

[0093] The result output module is used to output the risk information of myopia occurrence for the target individual within the preset future time period.

[0094] This invention also provides an electronic device, which can vary significantly due to differences in configuration or performance. It may include one or more central processing units (CPUs) (e.g., one or more processors) and memory, and one or more storage media (e.g., one or more mass storage devices) for storing applications or data. The memory and storage media can be temporary or persistent storage. The program stored in the storage media may include one or more modules, each module including a series of instruction operations on the electronic device. Furthermore, the processor may be configured to communicate with the storage media and execute the series of instruction operations stored in the storage media on the electronic device.

[0095] The electronic device may also include one or more power supplies, one or more wired or wireless network interfaces, one or more input / output interfaces, and / or one or more operating systems, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will understand that the electronic device structure in this embodiment does not constitute a limitation on the electronic device itself, and may include more or fewer components, or combinations of certain components, or different component arrangements.

[0096] This invention provides an electronic device structure that can vary significantly depending on configuration or performance. It may include one or more central processing units (CPUs) (e.g., one or more processors) and memory, and one or more storage media (e.g., one or more mass storage devices) for storing applications or data. The memory and storage media can be temporary or persistent storage. The program stored in the storage media may include one or more modules, each module including a series of instruction operations on the electronic device. Furthermore, the processor may be configured to communicate with the storage media and execute the series of instruction operations stored in the storage media on the electronic device.

[0097] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to perform the steps of the aforementioned method.

[0098] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system, device, or unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0099] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0100] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting myopia risk based on a joint model of ocular biological parameters, characterized in that, Includes the following steps: Acquire at least one measurement of ocular bioparameters of the target individual, the ocular bioparameters including axial length and corneal curvature data for calculating the axial ratio; Based on the eye biometric data of the target individual, a pre-trained joint myopia risk prediction model is used to process the data to obtain myopia risk information for the target individual within a preset future time period; wherein, the joint myopia risk prediction model is constructed through the following steps: Based on at least one ocular bioparameter measurement data of the target individual, longitudinal sub-model processing and survival sub-model processing are performed simultaneously through the joint model. The longitudinal sub-model is used to represent the trajectory of the change in axial ratio over time for the target individual and historical groups, and the survival sub-model is used to represent the risk of myopia. The longitudinal sub-model and the survival sub-model are associated through shared random effects parameters. The dependent variable of the longitudinal sub-model is the axial ratio; The fixed effects of the longitudinal sub-model include: age, sex, BMI, baseline spherical lens, cylindrical lens, astigmatism axis, corneal curvature K1 / K2, school type, and place of origin. The random effects of the longitudinal sub-model include: individual random intercept and random slope; Output the risk information of myopia occurrence for the target individual within the preset future time period.

2. The method according to claim 1, characterized in that, The longitudinal sub-model is a mixed-effects longitudinal model, which uses the axial ratio as the longitudinal response variable and expresses the axial ratio as a function of time. The mixed-effects longitudinal model includes at least fixed-effects parameters to describe the trend of population-level changes and random-effects parameters to describe the differences between the target individual and the historical population.

3. The method according to claim 2, characterized in that, The random effects parameters include at least individual random intercept parameters and individual random slope parameters. The individual random intercept parameters are used to characterize the deviation of the target individual from the initial level of the axial ratio, and the individual random slope parameters are used to characterize the deviation of the target individual's axial ratio from the rate of change over time relative to the population mean rate of change.

4. The method according to claim 1, characterized in that, The survival sub-model is a time-event analysis model, in which the time of the first occurrence of myopia is used as the survival time variable, and whether or not myopia occurs is used as the event state variable. The risk function of the survival sub-model incorporates at least one covariate, which includes an axial ratio obtained from at least one ocular bioparameter measurement of the target individual or a derived parameter calculated from the axial ratio.

5. The method according to claim 1, characterized in that, The shared random effects parameters, as latent variables, participate in the parameter estimation process of both the longitudinal sub-model and the survival sub-model, enabling the individual difference information in the longitudinal sub-model used to describe the trajectory of axial ratio change over time to be synchronously reflected in the risk function in the survival sub-model used to describe the risk of myopia.

6. The method according to claim 1, characterized in that, The method further includes a data preprocessing step, which includes data integrity verification, outlier identification, and data format standardization. The outlier identification is used to determine whether the axial length or corneal curvature data exceeds a preset reasonable range. Data identified as outliers will not be included in the calculation of the subsequent myopia risk joint prediction model. The data format standardization process is used to convert the eye biometric measurement data of the target individual into a data structure format consistent with the training phase of the joint myopia risk prediction model.

7. The method according to claim 1, characterized in that, In the step of obtaining at least one eye biometric measurement data of the target individual, when the target individual has only one eye biometric measurement data, the myopia risk joint prediction model estimates the trajectory of the target individual's axial ratio change over time based on the longitudinal distribution characteristics of the eye biometrics of the historical group and the eye biometric measurement data of the target individual, through the posterior distribution of the preset random effects parameter in the longitudinal sub-model.

8. A myopia risk prediction system based on a joint model of eye biological parameters, characterized in that, The system includes: The data acquisition module is used to acquire at least one measurement of ocular bioparameters of the target individual, the ocular bioparameters including axial length and corneal curvature data for calculating the axial ratio; The model processing module is used to process the eye bioparameter measurement data of the target individual by calling a pre-trained joint prediction model for myopia risk, so as to obtain the myopia risk information of the target individual in the future within a preset time period; wherein, the joint prediction model for myopia risk is constructed in the following way: Based on at least one ocular biometric measurement data of the target individual, longitudinal sub-model processing and survival sub-model processing are performed simultaneously through a joint model. The longitudinal sub-model is used to characterize the trajectory of the change in axial ratio over time in the data of the target individual and historical groups, and the survival sub-model is used to characterize the risk of myopia. The longitudinal sub-model and the survival sub-model are associated through shared random effects parameters. The dependent variable of the longitudinal sub-model is the axial ratio; The fixed effects of the longitudinal sub-model include: age, sex, BMI, baseline spherical lens, cylindrical lens, astigmatism axis, corneal curvature K1 / K2, school type, and place of origin. The random effects of the longitudinal sub-model include: individual random intercept and random slope; The result output module is used to output the risk information of myopia occurrence for the target individual within the preset future time period.

9. An electronic device comprising a memory and at least one processor, wherein the memory stores instructions; The at least one processor invokes the instructions in the memory to cause the electronic device to perform the steps of the myopia risk prediction method based on a joint model of eye bioparameters as described in any one of claims 1-7.

10. A computer-readable storage medium storing instructions thereon, characterized in that, When the instructions are executed by the processor, they implement the various steps of the myopia risk prediction method based on a joint model of eye bioparameters as described in any one of claims 1-7.

Images