CN108281197A - A method of relationship between analysis environmental factor and juvenile shortsightedness - Google Patents

Abstract

The invention discloses a method for analyzing the relationship between environmental factors and juvenile myopia. 3. The user's close-range work and outdoor exposure data and the user's eye parameters are respectively transmitted to the cloud platform for storage; step 4, the big data analysis system is used to analyze the close-range work and outdoor exposure data stored in the cloud platform database and the user's eye parameters. The eye parameters are denoised and converted according to the requirements of big data analysis, and then the user's eye behavior features are extracted, and the eye behavior features and the user's eye parameters are correlated. The method can monitor the user's real eye-use behavior and visual environment in real time, dynamically and objectively during use, and overcomes the problem in the prior art that the relationship between environmental factors and juvenile myopia cannot be accurately analyzed.

Description

A method to analyze the relationship between environmental factors and myopia in adolescents

technical field

The invention relates to the field of analyzing the relationship between myopia and the environment, in particular to a method for analyzing the relationship between environmental factors and the myopia of young people.

Background technique

The prevalence of myopia is increasing rapidly, and there is currently no effective prevention and control method. It is generally believed that myopia is the result of the combined effects of genes and environmental factors, and the rapid increase in the prevalence of myopia in the past few decades suggests that environmental factors are the leading cause. Environmental factors mainly include close working and outdoor exposure. So far, the specific role of environmental factors in the development of myopia is still unclear. The reason is that previous studies on the relationship between environmental factors and myopia were mainly conducted through questionnaires.

At present, some studies have used wearable devices that can objectively monitor close-range work or outdoor exposure to evaluate certain aspects of environmental factors, but no research has applied smart wearable devices that can objectively monitor both close-range work and outdoor exposure. equipment. For example, Leung et al. applied a head-mounted proximity work analyzer to objectively monitor users' close work conditions. These researchers used portable photoreceptors, Actiwatch, and FitSight fitnesstracker to monitor users' outdoor exposure, respectively.

There are two main disadvantages of the questionnaire method. First, it requires the respondents to answer the questions with a wide range of contents and specific content in the questionnaire according to their memory, which cannot avoid the recall bias, resulting in the inaccuracy of the statistical data. Second, it can only provide the aggregate data of environmental factors, but cannot record the high-density behavioral data of the respondents in real time and dynamically, so it cannot describe their behavior patterns (such as continuous work when working at close range, Or intermittent work, etc.), animal experiments have confirmed that behavior patterns have an important impact on the development of diopter.

With the development of information technology, some of the methods mentioned above can objectively monitor environmental factors, but they also have some major disadvantages. First, these methods can only monitor certain aspects of environmental factors. For example, the head-mounted proximity analyzer can only monitor close-range work, and the Actiwatch can only monitor outdoor exposure time. This is obviously incomplete and cannot affect the environment. A comprehensive assessment of these factors cannot fully reveal the effect of environmental factors on the occurrence and development of myopia. Second, the ambient illuminance monitored by these methods is not the illuminance accepted by the eyes. Third, these methods do not have a complete data storage platform, and cannot obtain large-scale myopia-related behavioral data of adolescents, and cannot generate corresponding behavioral databases, let alone use big data to mine behavioral data and diopters. connection between.

Therefore, it provides a real-time, dynamic and objective monitoring of the user's real eye behavior and visual environment during use, which overcomes the shortcomings of previous methods that have recall bias and can only provide total environmental data; The two dimensions of the real working distance and the surrounding environment illuminance overcome the shortcomings of the previous method that the monitoring dimension is single and the measured data is not the real environment data of the eyes; the cloud platform with stored data overcomes the inability of the previous method to be stored on a large scale The shortcoming of data; And utilize big data method to data excavation and analysis, is expected to really determine the method for the quantitative relationship between environmental factor and myopia between a kind of analysis environmental factor and the relation between adolescent myopia is the present invention urgently needs to solve The problem.

Contents of the invention

In view of the above technical problems, the purpose of the present invention is to provide a real-time, dynamic and objective monitoring of the user's real eye behavior and visual environment during use, which overcomes the recall bias in the previous methods and can only provide total environmental data. It can monitor the two dimensions of the real working distance close to the eyes and the surrounding environment illumination at the same time, which overcomes the shortcomings of the previous method that the monitoring dimension is single and the measured data is not the real environment data of the eyes; it has a cloud platform for storing data, It overcomes the shortcomings of previous methods that cannot store data on a large scale; and uses big data methods to mine and analyze data, and it is expected to truly determine the quantitative relationship between environmental factors and myopia. An analysis of the relationship between environmental factors and adolescent myopia Methods.

In order to achieve the above object, the present invention provides a method for analyzing the relationship between environmental factors and juvenile myopia, said method comprising:

Step 1. Simultaneously objectively monitor and collect users’ close work and outdoor exposure data;

Step 2, obtaining user's eye parameters;

Step 3, the user's close work and outdoor exposure data and the user's eye parameters are respectively transmitted to the cloud platform for storage;

Step 4: Use the big data analysis system to perform noise reduction and conversion on the close-range work and outdoor exposure data stored in the cloud platform database and the user's eye parameters according to the requirements of big data analysis, and then extract the user's eye behavior characteristics , and correlate the eye behavior characteristics with the user's eye parameters, and then clarify the impact of environmental factors on the onset of myopia.

Preferably, the user's eye parameters include: objective diopter and eye axis data.

Preferably, the big data analysis system includes: a CPU processor, a data management module, a behavioral feature extraction module, an association model module and an optimal model decision module and an environmental impact index generation module; wherein the data management module is used for cloud platform The stored short-distance work and outdoor exposure data and the user's eye parameters are denoised and converted according to the requirements of big data analysis; the behavioral feature extraction module is used to utilize the close-distance work data and outdoor exposure data stored on the cloud platform to The user's eye behavior features are extracted to describe the behavior features of different users; the association model building module is used to use the association model to establish the relationship between the behavior features and the eye parameters after the user's eye behavior features are extracted. The connection is to establish the relationship between behavioral characteristics and diopter progress; the optimal model decision module is used to select and determine the optimal model; the environmental impact index generation module is used to generate a model that can reflect the user's eye behavior habits. The environmental impact index; the CPU processor is used to coordinate the work of each module and data analysis.

Preferably, the data management module uses fast Fourier transformation to filter out high-frequency data, that is, noise.

Preferably, the user's eye behavior features include: working distance (VD) and ambient illuminance (EI), and the noise-reduced VD data and EI data are mapped to a 2-dimensional space (VD-EI space), The vertical axis is VD, the horizontal axis is log10(EI), and the user's behavior distribution curve is obtained, and then the behavior curves of all users are superimposed to obtain the behavior distribution heat map of the user group, which immediately depicts the user Eye behavior characteristics of groups.

Preferably, the index of diopter progression is at least 2-year spherical equivalent power (SER) change value ΔSER and eye axis (AL) change value ΔAL; wherein, the SER=S+1/2C; S and C are the spherical and cylindrical powers obtained by the computer refractometer after ciliary muscle paralysis; the ΔSER=the last time of SER-SER baseline, ΔAL=the last time of AL-AL baseline; the last time of SER and the last time of AL respectively represent the user The diopter and axial data provided last time, the SER baseline and AL baseline respectively represent the diopter and axial data provided by the user for the first time.

Preferably, before the association model building module works, it is also necessary to explain the key assumptions for establishing the model, and then it is necessary to consider the influence of continuous pixels around it to obey the assumption: in units of pixels, calculate each user included in each pixel The proportion of the behavior time to the total time of the user group behavior (PoT), the radial basis function (RBF kernel function) is used to weight the pixels at different distances, and then the weighted linear regression (WLR) is used to analyze PoT and SER The relationship between the pixels closer to the pixel has greater influence, and the farther away the pixel influence is smaller; for the RBF kernel function of two pixels x and x', given the weight definition of x and x' relative to x for:

||x _ind -x′ _ind || ² is the square of the Euclidean distance between two pixels, the entire VD-EI space is divided into 40*40 pixels, and a pair of index values can be assigned to each pixel to calculate For the Euclidean distance between any two pixels, the RBF kernel function can assign small weights to far pixels and large weights to near pixels.

Preferably, when analyzing a certain pixel, it is necessary to define an area of a certain size with the pixel as the center of the circle to determine the scope of influence, where the size of this area is defined by the parameter δ (0≤δ≤20), and twice the δ is If a certain pixel is the radius of a circle whose center is the circle, then all pixels in the circle have a certain influence on this pixel.

Preferably, each pixel can establish a weighted linear regression model between PoT and diopter progression, the slope (K) of the model represents the nature and strength of the association between a certain behavioral feature and diopter progression, if the value is positive , indicating that the eye habits are good, the larger the positive value, the better the behavior habits, which means that the user’s eye behavior has a protective effect on myopia as a whole; the negative value indicates that the eye habits are bad, and the smaller the negative value is, the behavior habits The worse it is, the user's eye use behavior has a dangerous effect on myopia as a whole.

Preferably, the smart wearable device is used to objectively monitor and collect the user's close-range work and outdoor exposure data at the same time, and the frequency of collecting data is 4-6s/time, and the data collected by the smart wearable device will pass through Send it to the mobile APP, and the APP then transmits the data to the server through the network.

According to the above-mentioned technical solution, a method for analyzing the relationship between environmental factors and juvenile myopia provided by the present invention can monitor the user's real eye-use behavior and visual environment in real time, dynamically and objectively during use, and overcomes the existing problems of previous methods. The shortcomings of recall bias and can only provide the total amount of environmental data; it can simultaneously monitor the two dimensions of the real working distance close to the eyes and the surrounding environment illuminance, which overcomes the single dimension of monitoring in the previous method and the measured data is not the real environment data of the eyes The shortcomings of the data; the cloud platform for storing data overcomes the shortcomings of previous methods that cannot store data on a large scale; and the use of big data methods to mine and analyze data is expected to truly determine the quantitative relationship between environmental factors and myopia.

Other features and advantages of the present invention will be described in detail in the following detailed description.

Description of drawings

The accompanying drawings are used to provide a further understanding of the present invention, and constitute a part of the description, together with the following specific embodiments, are used to explain the present invention, but do not constitute a limitation to the present invention. In the attached picture:

Fig. 1 is a flow block diagram of a method for analyzing the relationship between environmental factors and juvenile myopia provided under a preferred embodiment of the present invention;

Fig. 2 is a schematic structural diagram of a smart wearable device in a method for analyzing the relationship between environmental factors and juvenile myopia provided under a preferred embodiment of the present invention;

Fig. 3 is an assembly diagram of a smart wearable device worn on glasses in a method for analyzing the relationship between environmental factors and juvenile myopia provided by a preferred embodiment of the present invention.

Explanation of reference signs

1 UV sensor 2 Bluetooth

3 Three-axis angular velocity sensor 4 Distance sensor

5 light intensity sensor

Detailed ways

Specific embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings. It should be understood that the specific embodiments described here are only used to illustrate and explain the present invention, and are not intended to limit the present invention.

As shown in Figures 1-3, the present invention provides a method for analyzing the relationship between environmental factors and juvenile myopia, characterized in that the method includes: Step 1, the user's close work and outdoor exposure The data is objectively monitored and collected at the same time; step 2, obtain the user's eye parameters; step 3, transmit the user's close work and outdoor exposure data and the user's eye parameters to the cloud platform for storage; step 4, use the large The data analysis system performs noise reduction and conversion on the close-range work and outdoor exposure data stored in the cloud platform database and the user's eye parameters according to the requirements of big data analysis, and then extracts the characteristics of the user's eye-use behavior, and analyzes the user's eye parameters. Behavioral characteristics and user's eye parameters are correlated to clarify the impact of environmental factors on the onset of myopia.

According to the above-mentioned technical solution, a method for analyzing the relationship between environmental factors and juvenile myopia provided by the present invention can monitor the user's real eye-use behavior and visual environment in real time, dynamically and objectively during use, and overcomes the existing problems of previous methods. The shortcomings of recall bias and can only provide the total amount of environmental data; it can simultaneously monitor the two dimensions of the real working distance close to the eyes and the surrounding environment illuminance, which overcomes the single dimension of monitoring in the previous method and the measured data is not the real environment data of the eyes The shortcomings of the data; the cloud platform for storing data overcomes the shortcomings of previous methods that cannot store data on a large scale; and the use of big data methods to mine and analyze data is expected to truly determine the quantitative relationship between environmental factors and myopia.

In a preferred embodiment of the present invention, the user's eye parameters include: objective diopter and eye axis data, these two parameters are the user's main eye parameters, of course the user's eye parameters in the present invention Not limited to these two ocular parameters.

In a preferred embodiment of the present invention, the big data analysis system includes: a CPU processor, a data management module, a behavior feature extraction module, an association model module, an optimal model decision module and an environmental impact index generation module; wherein, The data management module is used for noise reduction and conversion of the close-range work and outdoor exposure data stored in the cloud platform and the user's eye parameters according to the requirements of big data analysis; Extract the user's eye behavior characteristics from the distance work data and outdoor exposure data, and describe the behavior characteristics of different users; the association model building module is used to use the association model to establish The connection between behavioral characteristics and eye parameters is to establish the relationship between behavioral characteristics and diopter progression; the optimal model decision module is used to select and determine the optimal model; the environmental impact index generation module is used to generate The environmental impact index that reflects the user's eye behavior habit; the CPU processor is used for coordinating the work of various modules and data analysis.

In a preferred embodiment of the present invention, the data management module uses fast Fourier transformation to filter out high-frequency data, that is, noise. The behavioral data obtained from the cloud platform is presented in the order of the collection time, and a piece of behavioral data is displayed every 4-6s. Each data includes the time point of data collection and the working distance corresponding to the time point. And the surrounding environment illuminance, this form of raw data, the distribution characteristics of the data cannot be known, and certain processing is required. Taking a user as an example, number his behavioral data in Arabic numerals in time series, use the number as the abscissa, and the environmental illumination (EI) or working distance (VD) corresponding to the number as the ordinate, to get the user's Working distance distribution map or surrounding environment illuminance distribution map. From the data distribution diagram, it can be seen that the noise of the data is too large for big data analysis, and the data must be denoised. Fast Fourier transform (FFT) is an effective noise reduction method, which filters out high-frequency data (noise), and the data after noise reduction has a distribution that is more suitable for analysis.

In a preferred embodiment of the present invention, the user's eye behavior characteristics include: working distance (VD) and illuminance (EI) of the surrounding environment, and the VD data and EI data after noise reduction are mapped to a 2 dimensional space (VD-EI space), the vertical axis is VD, and the horizontal axis is log10(EI), then the user's behavior distribution curve is obtained, and then the behavior curves of all users are superimposed to obtain the behavior distribution of the user group The heat map immediately depicts the eye-using behavior characteristics of this user group. Behavioral data consists of 3 features: continuous time series, working distance or illuminance corresponding to each collection time point. The present invention focuses on the user's eye-using behavior characteristics, so the time dimension can be ignored, and the remaining two characteristics are analyzed: the working distance (VD) and the illuminance (EI) of the surrounding environment. Taking a certain user as an example, map the denoised VD and EI data to a 2-dimensional space (VD-EI space), the vertical axis is VD, and the horizontal axis is log10(EI), then the behavior of a certain user can be obtained The distribution curve graph immediately depicts the user's eye-using behavior characteristics. Then superimpose the behavior curves of all users separately to obtain the behavior distribution heat map of the user group, which immediately depicts the eye-using behavior characteristics of the user group.

In a preferred embodiment of the present invention, the index of diopter progression is at least 2 years of spherical equivalent power (SER) change value ΔSER and eye axis (AL) change value ΔAL; wherein, the SER=S+1/2C; S and C are the spherical power and cylindrical power obtained by the computer refractometer after ciliary muscle paralysis; the ΔSER=the last time of SER-SER baseline, ΔAL=the last time of AL-AL baseline ; SER last time and AL last time respectively represent the last diopter and eye axis data provided by the user, SER baseline and AL baseline respectively represent the diopter and eye axis data provided by the user for the first time. After obtaining the total behavior distribution heat map of the user group through the behavior feature extraction module, the overall eye-using behavior characteristics of the user group can be obtained. Obviously, the overall eye behavior characteristics cannot be correlated with the diopter progression. Therefore, the overall behavior distribution heat map is divided into 40x40 grids (pixels), and each pixel represents a local behavior characteristic. Next, a correlation between local behavioral features and refraction progression needs to be established.

In a preferred embodiment of the present invention, before establishing the association model, it is necessary to explain the key assumptions for establishing the model. The key assumption is that people's eye behavior is spatially continuous, which means that a person's eye behavior covers A slice of continuous pixels. Therefore, when analyzing each pixel, the influence of continuous pixels around it should be considered to obey the assumption.

In pixels, calculate the ratio of the time of each subject's behavior contained in each pixel to the total time of these subjects' behaviors (PoT). For a subject, PoT is its The ratio of the time spent to the time spent in all pixels that the behavior curve passes through reflects the proportion of time spent by the subject in this pixel (behavior feature). So far, there is an independent variable PoT linked to the behavioral characteristics of the subjects, and there are also dependent variables Δ _SER and Δ _AL that reflect the progress of the subjects’ diopter, so that the relationship between the subjects’ PoT per unit pixel and their Association relationship between diopter progression. As mentioned earlier, when analyzing a pixel, the influence of surrounding continuous pixels needs to be considered. Obviously, the closer the pixel is to the pixel, the greater the impact, and the farther the pixel is, the less impact, so the radial basis function (RBF kernel function) is used to weight the pixels at different distances, and then the weighted linear regression (WLR ) to analyze the relationship between PoT and SER;

For the RBF kernel function of 2 pixels x and x', the relative weight of given x and x' is defined as:

||x _ind -x′ _ind || ² is the square of the Euclidean distance between two pixels, the entire VD-EI space is divided into 40*40 pixels, and a pair of index values can be assigned to each pixel to calculate For the Euclidean distance between any two pixels, the RBF kernel function can assign small weights to far pixels and large weights to near pixels.

In a preferred embodiment of the present invention, when analyzing a certain pixel, since the surrounding pixels have a certain influence, it is necessary to define an area of a certain size with the pixel as the center of the circle to determine the scope of influence, where the parameter δ( 0≤δ≤20) defines the size of this area, 2 times δ is the radius of a circle centered on a certain pixel, then all pixels in the circle have a certain influence on the pixel;

Among them, the optimal model determination module: this module is mainly used to determine which correlation model is the optimal model. Different values of δ will result in different relationship models between PoT and diopter progression, so it is necessary to verify which value of δ is the most accurate relationship model established. If there is no behavior distribution in a pixel and the pixels contained in the circle with its center as the radius of 2δ, that is, the PoT is 0, then the pixel cannot establish a weighted linear regression model between the PoT and the diopter progression. At this time, it can be considered The slope (K) of the model is 0. In extreme cases, if there is no behavior distribution in all pixels and the pixels contained in the circle with a radius of 2δ as the center of the circle, the relationship model between PoT and diopter progression cannot be established. Therefore, the fewer the number of pixels with K=0, the more effectively the relationship between PoT and diopter progression can be reflected. Obviously, the number of pixels with K=0 has a great relationship with the value of δ. The larger the value of δ, the larger the area of influence around a certain pixel, and the more pixels are contained in the area, and there is no behavioral distribution at all. The smaller the probability, the more likely it is to model the relationship between PoT and diopter progression (ie, the smaller the probability of K=0). Therefore, theoretically speaking, the larger the value of δ, the more accurate the established relationship model between PoT and diopter progression.

In a preferred embodiment of the present invention, each pixel can establish a weighted linear regression model between PoT and diopter progression, and the slope (K) of the model represents the correlation between a certain behavioral feature and diopter progression and intensity, if the value is positive, it means that the eye habits are good, the greater the positive value, the better the behavior habits, which means that the user’s eye behavior has a protective effect on myopia as a whole; the value is negative, indicating that the eyes Bad habits, the smaller the negative value, the worse the behavior habits, which means that the user's eye use behavior has a dangerous effect on myopia as a whole; wherein, the environmental impact index generation module included in the big data analysis system is used to generate The environmental impact index that can reflect the user's eye behavior habits: In this invention, the parameter is named "environmental impact index" as a custom, just like anemia, it can be determined according to the content of hemoglobin; high blood pressure, according to The values of systolic blood pressure and diastolic blood pressure can be determined the same; the present invention hopes to establish a parameter that can reflect the quality of eye-using behavior habits: "environmental impact index".

In a preferred embodiment of the present invention, the smart wearable device is used to objectively monitor and collect the user's close-range work and outdoor exposure data at the same time, and the frequency of collecting data is 4-6s/time. The data collected by the wearable device will be sent to the mobile phone APP through Bluetooth, and the APP will then transmit the data to the server through the network.

The preferred embodiment of the present invention has been described in detail above in conjunction with the accompanying drawings, but the present invention is not limited to the specific details of the above embodiment, within the scope of the technical concept of the present invention, various simple modifications can be made to the technical solution of the present invention, These simple modifications all belong to the protection scope of the present invention.

In addition, it should be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable way if there is no contradiction. The combination method will not be described separately.

In addition, various combinations of different embodiments of the present invention can also be combined arbitrarily, as long as they do not violate the idea of the present invention, they should also be regarded as the disclosed content of the present invention.

Claims (10)

1. a method for analyzing the relationship between environmental factors and juvenile myopia, is characterized in that, described method comprises: Step 1. Simultaneously objectively monitor and collect users’ close work and outdoor exposure data; Step 2, obtaining user's eye parameters; Step 3, the user's close work and outdoor exposure data and the user's eye parameters are respectively transmitted to the cloud platform for storage; Step 4: Use the big data analysis system to perform noise reduction and conversion on the close-range work and outdoor exposure data stored in the cloud platform database and the user's eye parameters according to the requirements of big data analysis, and then extract the user's eye behavior characteristics , and correlate the eye behavior characteristics with the user's eye parameters, and then clarify the impact of environmental factors on the onset of myopia. 2. The method for analyzing the relationship between environmental factors and juvenile myopia according to claim 1, wherein the user's eye parameters include: objective diopter and eye axis data. 3. the method for analyzing the relationship between environmental factors and juvenile myopia according to claim 1, is characterized in that, big data analysis system comprises: CPU processor, data management module, behavior feature extraction module, association model module and the most Optimal model determination module and environmental impact index generation module; among them, The data management module is used to perform noise reduction and conversion on the close-range work and outdoor exposure data stored on the cloud platform and the user's eye parameters according to the requirements of big data analysis; The behavioral feature extraction module is used to extract the user's eye-using behavioral features using the close-range work data and outdoor exposure data stored on the cloud platform, and describe the behavioral features of different users; The association model building module is used to use the association model to establish the connection between the behavior characteristics and the eye parameters after the extraction of the user's eye behavior characteristics is completed, that is, to establish the association between the behavior characteristics and the diopter progress; The optimal model determination module is used to select and determine the optimal model; The environmental impact index generation module is used to generate an environmental impact index that can reflect the user's eye-using behavior habits; The CPU processor is used for coordinating the work of each module and data analysis. 4. the method for analyzing the relationship between environmental factors and juvenile myopia according to claim 3 is characterized in that, the data management module utilizes fast Fourier transform to filter out high-frequency data, that is, noise. 5. the method for analyzing the relationship between environmental factors and juvenile myopia according to claim 3, is characterized in that, the eye-using behavior characteristic of described user comprises: the illuminance (EI) of working distance (VD) and surrounding environment, Map the denoised VD data and EI data to a 2-dimensional space (VD-EI space), the vertical axis is VD, and the horizontal axis is log10(EI), then the user's behavior distribution curve is obtained, and then all users' The behavior curves are superimposed separately to obtain the behavior distribution heat map of the user group, which immediately depicts the eye-using behavior characteristics of the user group. 6. the method for analyzing the relationship between environmental factors and juvenile myopia according to claim 3, is characterized in that, the index of described diopter progress is at least the change value ΔSER of the spherical equivalent power (SER) of 2 years and The change value ΔAL of the eye axis (AL); where, The SER=S+1/2C; S and C are respectively the spherical power and cylinder power obtained by the computer optometry after the ciliary muscle is paralyzed; The ΔSER=the last time of SER-SER baseline, ΔAL=the last time of AL-AL baseline; diopter and eye axis data. 7. the method for the relation between analysis environmental factor and juvenile myopia according to claim 6, is characterized in that, also need to explain the key hypothesis of setting up this model before described association model building module work, then need to consider its surrounding The influence of continuous pixels is subject to the assumption: in units of pixels, calculate the ratio of the time of each user's behavior contained in each pixel to the total time of the user group's behavior (PoT), and use the radial basis function (RBF kernel function) to give Pixels at different distances are weighted, and then weighted linear regression (WLR) is used to analyze the relationship between PoT and SER. The closer the pixel is to the pixel, the greater the impact, and the farther the pixel is, the smaller the impact; For the RBF kernel function of 2 pixels x and x', the weight of given x and x' relative to x is defined as: ||x _ind -x′ _ind || ² is the square of the Euclidean distance between two pixels, the entire VD-EI space is divided into 40*40 pixels, and a pair of index values can be assigned to each pixel to calculate For the Euclidean distance between any two pixels, the RBF kernel function can assign small weights to far pixels and large weights to near pixels. 8. the method for analyzing the relationship between environmental factors and juvenile myopia according to claim 7, is characterized in that, when analyzing a certain pixel, it is necessary to define the area of a certain size with the pixel as the center of circle to determine the scope of influence, Among them, the size of this area is defined by the parameter δ (0≤δ≤20), and the double δ is the radius of a circle with a certain pixel as the center, so all the pixels in the circle have a certain influence on the pixel. 9. the method for analyzing the relationship between environmental factors and juvenile myopia according to claim 3, is characterized in that, each pixel can set up the weighted linear regression model between PoT and diopter progress, the slope of this model ( K) represents the nature and strength of the correlation between a certain behavioral feature and the diopter progression. If the value is positive, it means that the eye habits are good. The larger the positive value, the better the behavior habits, which means that the user's eye behavior is generally good for myopia. Eyes have a protective effect; a negative value indicates poor eye habits, and the smaller the negative value, the worse the behavior, which means that the user's eye behavior as a whole has a dangerous effect on myopia. 10. the method for analyzing the relationship between environmental factors and juvenile myopia according to claim 1, is characterized in that, utilizes smart wearable device to carry out objective monitoring and collection simultaneously to user's short-distance work and outdoor exposure data, wherein The frequency of data collection is 4-6s/time, and the data collected by the smart wearable device will be sent to the mobile phone APP through Bluetooth, and the APP will then transmit the data to the server through the network.