Data processing method and system based on cataract detection

Abstract

The invention provides a data processing method and system based on cataract detection, and the method comprises the steps: obtaining the historical data of different dimensions of a user, and carrying out the training of an AI decision model according to the historical data of different dimensions; acquiring to-be-analyzed data, inputting the to-be-analyzed data into the trained AI decision model, and outputting an analysis result; wherein in the process of training the AI decision model, the entropy flow roaming algorithm is adopted to continuously adjust weights of data of different dimensions so as to optimize the performance of the model, and specifically, the entropy flow roaming algorithm optimizes weight distribution through an entropy flow potential function and a dynamic balance mechanism, can adaptively adjust a search range, is more suitable for complex characteristics of medical data, and improves the accuracy of the medical data. And finally, the accuracy and reliability of a data processing result based on cataract detection are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of data processing based on cataract detection, and particularly relates to a data processing method and system based on cataract detection. Background Art

[0002] Cataract is one of the main causes of blindness globally, especially with a relatively high incidence in the elderly population. Existing clinical cataract detections rely on professional ophthalmologists to manually interpret medical data such as slit lamp images and fundus images. This traditional data processing method has two major technical bottlenecks: one is the low efficiency of manual annotation, which takes 5 - 10 minutes to annotate a single sample; the other is limited by the uneven distribution of medical resources.

[0003] With the application of artificial intelligence technology in the field of medical imaging, it has become possible to use AI technology for disease screening. However, existing AI technologies have deficiencies in data utilization and decision-making accuracy, especially in the weight optimization link.

[0004] The existing weight optimization link mainly relies on temperature parameters to control the search process. Its convergence speed and global search ability are limited by the fixed cooling strategy, and it is difficult to achieve precise weight allocation in the high-dimensional and small-sample scenarios of medical data. Additionally, there may be problems such as premature convergence and high computational complexity, which also pose challenges when dealing with complex medical data. All of these lead to poor cataract detection effects. Summary of the Invention

[0005] Based on this, embodiments of the present invention provide a data processing method and system based on cataract detection, aiming to improve the accuracy and reliability of data processing results based on cataract detection by integrating multi-dimensional data and using a unique entropy flow roaming algorithm to determine the weights of various data types in AI decision-making.

[0006] The first aspect of the embodiments of the present invention provides a data processing method based on cataract detection, and the method includes: Obtain historical data of different dimensions for a user to perform cataract detection, and train an AI decision model according to the historical data of different dimensions. The historical data includes at least user basic data and eye data related to cataract detection; Obtain data to be analyzed, input the data to be analyzed into the trained AI decision model, and output an analysis result; Wherein, during the process of training the AI decision model, the entropy flow roaming algorithm is used to continuously adjust the weights of data of different dimensions to optimize the performance of the model. Specifically, preprocess the historical data of different dimensions, and perform few-sample correlation analysis on the preprocessed data to obtain a correlation analysis result; According to the correlation analysis results, assign initial weights to data of different dimensions; According to the initial weights, through simulating the adaptive switching mechanism from high entropy state to low entropy state, achieve global optimal search in the solution space, and output the optimal weight vectors of data of different dimensions.

[0007] Further, in the step of performing few-sample correlation analysis on the preprocessed data to obtain the correlation analysis results, the calculation formula for few-sample correlation analysis is:

[0008] where d i is the rank difference between the data type and the analysis result in the i-th sample, n is the number of samples, and r s is the correlation analysis result.

[0009] Further, in the step of assigning initial weights to data of different dimensions according to the correlation analysis results, the initial weights are expressed as:

[0010] where is the initial weight of the j-th type of data, r sj is the Spearman rank correlation coefficient between the j-th type of data and the cataract analysis result, r sk is the Spearman rank correlation coefficient between the k-th type of data and the cataract analysis result, and m is the total number of data types.

[0011] Further, the step of achieving global optimal search in the solution space according to the initial weights through simulating the adaptive switching mechanism from high entropy state to low entropy state and outputting the optimal weight vectors of data of different dimensions includes: Define the entropy flow potential function and set the initial entropy flow intensity; Generate a perturbation term through a standard normal distribution random number, calculate a new weight vector in combination with the current entropy flow intensity, and perform normalization processing to ensure that the sum of the weights is 1; Calculate the entropy difference between the current state and the perturbed state, and obtain different solutions according to the entropy difference result; Control the dynamic entropy decay process to satisfy the exponential decay characteristic to ensure convergence within a finite number of steps; When the entropy flow intensity is less than the termination entropy value, output the optimal weight vector.

[0012] Further, in the step of generating a perturbation term through a standard normal distribution random number and calculating a new weight vector in combination with the current entropy flow intensity, the perturbation formula is:

[0013] Among them, w' is the new weight, w is the current weight, β is the perturbation coefficient, and S k is the current entropy flow intensity, and randn() is a function for generating random numbers with a standard normal distribution.

[0014] Furthermore, in the step of calculating the entropy difference between the current state and the perturbed state and obtaining different solutions according to the entropy difference result, the calculation formula for the entropy difference is:

[0015] Among them, ΔE is the entropy difference, E(w') is the entropy flow energy of the perturbed state, and E(w) is the entropy flow energy of the current state; When , accept the new solution; When , with a probability of , accept the new solution.

[0016] The second aspect of the embodiments of the present invention provides a data processing system based on cataract detection, which is used to implement the data processing method based on cataract detection provided in the first aspect. The system includes: A training module, configured to obtain historical data of different dimensions of a user, and train an AI decision model according to the historical data of different dimensions; An input module, configured to obtain data to be analyzed, input the data to be analyzed into the trained AI decision model, and output an analysis result; Among them, in the process of training the AI decision model, the entropy flow roaming algorithm is used to continuously adjust the weights of data in different dimensions to optimize the performance of the model. Specifically, the historical data in different dimensions is preprocessed, and the preprocessed data is subjected to few-sample correlation analysis to obtain a correlation analysis result; According to the correlation analysis result, initial weights are assigned to data in different dimensions; According to the initial weights, through a simulated adaptive switching mechanism from a high entropy state to a low entropy state, global optimal search is realized in the solution space, and an optimal weight vector of data in different dimensions is output.

[0017] The third aspect of the embodiments of the present invention provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the data processing method based on cataract detection provided in the first aspect.

[0018] The fourth aspect of the embodiments of the present invention provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the data processing method based on cataract detection provided in the first aspect.

[0019] A data processing method and system based on cataract detection provided in an embodiment of the present invention. The method obtains historical data of different dimensions of a user, and trains an AI decision-making model according to the historical data of different dimensions; obtains data to be analyzed, inputs the data to be analyzed into the trained AI decision-making model, and outputs an analysis result; wherein, in the process of training the AI decision-making model, the entropy flow roaming algorithm is used to continuously adjust the weights of data of different dimensions to optimize the performance of the model. Specifically, the entropy flow roaming algorithm optimizes the weight allocation through an entropy flow potential function and a dynamic balance mechanism, can adaptively adjust the search range, is more suitable for the complex characteristics of medical data, and finally improves the accuracy and reliability of the data processing result based on cataract detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 FIG. is a flowchart of an implementation of a data processing method based on cataract detection provided in Embodiment 1 of the present invention; Figure 2 FIG. is a structural block diagram of a data processing system based on cataract detection provided in Embodiment 2 of the present invention; Figure 3 FIG. is a structural block diagram of an electronic device provided in Embodiment 3 of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0021] To facilitate the understanding of the present invention, the present invention will be described more comprehensively below with reference to the relevant drawings. Several embodiments of the present invention are shown in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, these embodiments are provided so that the disclosure of the present invention will be thorough and comprehensive.

[0022] It should be noted that when an element is referred to as being "fixedly provided on" another element, it can be directly on the other element or there can also be an intermediate element. When an element is considered to be "connected" to another element, it can be directly connected to the other element or there may be an intermediate element at the same time. The terms "vertical", "horizontal", "left", "right" and similar expressions used herein are only for the purpose of illustration.

[0023] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs. The terms used in the description of the present invention herein are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0024] Embodiment 1 According to an embodiment of the present invention, a data processing method based on cataract detection is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0025] In the first embodiment, a data processing method based on cataract detection is provided, which can be used in an electronic device, such as a computer. Please refer to Figure 1 , Figure 1 FIG. shows an implementation flowchart of a data processing method based on cataract detection provided in the first embodiment of the present invention, which specifically includes steps S01 to S02.

[0026] Step S01, obtain historical data of different dimensions for the user to perform cataract detection, and train the AI decision model according to the historical data of different dimensions. The historical data at least includes user basic data and eye data related to cataract detection.

[0027] Specifically, the obtained historical data of different dimensions includes high-risk factors such as gender, age, eye image data, binocular vision data, visual sensitivity data, presence or absence of surgical indications, presence or absence of surgical contraindications, medical history (presence or absence of diabetes, presence or absence of fundus diseases, presence or absence of refractive errors, etc.). Among them, the eye data related to cataract detection is eye image data, binocular vision data, and visual sensitivity data, and other data is user basic data. It can be understood that gender is a dimension of data, age is another dimension of data. In addition, a dimension can also be called a type, that is, gender is a type of data, and age is another type of data.

[0028] In this embodiment, the AI decision model is a multi-modal weighted fusion model using a voting mechanism, and different weights are assigned to various types of data according to their influence on the cataract analysis result.

[0029] Step S02, obtain the data to be analyzed, input the data to be analyzed into the trained AI decision model, and output the analysis result.

[0030] Among them, in the process of training the AI decision model, the entropy flow roaming algorithm is used to continuously adjust the weights of data of different dimensions to optimize the performance of the model. Specifically, the entropy flow roaming algorithm optimizes the weight allocation through the entropy flow potential function and the dynamic balance mechanism.

[0031] It should be noted that the step of continuously adjusting the weights of data of different dimensions by using the entropy flow roaming algorithm includes: Preprocess historical data of different dimensions, and perform few-sample correlation analysis on the preprocessed data to obtain the correlation analysis results. Specifically, first preprocess the historical data of different dimensions, including operations such as data cleaning and normalization, to ensure the quality and consistency of the data and make different types of data comparable. In the data cleaning process, remove obviously incorrect or abnormal data records by setting reasonable data thresholds and outlier detection rules. For example, if the age data shows negative values or values far beyond the normal human lifespan range, it is regarded as an outlier for processing. For eye image data, use image enhancement technology to improve the clarity and contrast of the images for better extraction of image features. For age data, use the normalization formula , and transform it into the 0-1 interval, where x is the original age value, and xmin and xmax are the minimum and maximum values in the age data respectively, which is convenient for subsequent calculations.

[0032] Furthermore, since the actually collected data may have a limited sample size, use the Spearman rank correlation coefficient applicable to few samples to analyze the correlation between various types of data and the actual diagnosis results of cataracts. The calculation formula for the few-sample correlation analysis is:

[0033] where di is the rank difference between the data type and the analysis result in the i-th sample, n is the number of samples, and rs is the correlation analysis result. By using the calculation formula for the few-sample correlation analysis, calculate the correlation coefficients between various types of data such as gender, age, and eye image data and the cataract analysis results respectively, so as to quantify the degree of association between them. For example, age growth is usually positively correlated with the incidence probability of cataracts. The correlation coefficient between age and the cataract analysis result calculated by this formula can intuitively reflect the tightness of this association. It should be noted that the Spearman rank correlation coefficient, that is, the correlation analysis result, is used to measure the monotonic correlation between two variables (linear relationship is not required), and is applicable to non-normal distribution or ranked data. In the embodiments of the present invention, the original data is converted into ranks, and the correlation is analyzed by calculating the difference between the two sets of ranks. Exemplarily, through the expert evaluation method, the user's age can be divided into ranks. 80 years old is rank 5, 72 years old is rank 4, 65 years old is rank 3, 60 years old is rank 2, and 58 years old is rank 1. The cataract analysis results are divided into ranks and also divided into 5 ranks according to different severity levels; According to the correlation analysis results, assign initial weights to the data of different dimensions. Specifically, the initial weights are expressed as:

[0034] where, is the initial weight of the j-th type of data, r sj is the Spearman rank correlation coefficient between the j-th type of data and the cataract analysis result, r sk is the Spearman rank correlation coefficient between the k-th type of data and the cataract analysis result, m is the total number of data types. It should be noted that the purpose of this formula is to normalize the correlation coefficients of various types of data so that the sum of the weights of all data types is 1, ensuring the rationality and additivity of weight allocation; According to the initial weight, through the adaptive switching mechanism that simulates the transition from a high-entropy state to a low-entropy state, global optimal search is realized in the solution space, and the optimal weight vector of data in different dimensions is output. Specifically, first, the entropy flow potential function is defined and the initial entropy flow intensity is set. It should be noted that the entropy flow potential function is defined as: F(w)=-E(w), where w is the current weight, and E(w) represents the entropy flow energy of the system, which is defined by the negative value of the objective function (screening accuracy), that is, E(w)=-Accuracy(w). According to the second law of thermodynamics, the entropy increase of an isolated system satisfies . In EFW, the entropy flow potential function F(w)=-E(w) simulates the process of minimizing the system energy, and drives the optimization direction through negative entropy to ensure that the algorithm converges to a high-precision solution. Subsequently, the initial entropy flow intensity is set , with a value range of 50-200, preferably 100 (to control the range of random perturbations); perturbation coefficient , with a value range of 0.05-0.2, preferably 0.1; entropy decay rate , with a value range of 0.9-0.99, preferably 0.95 (determining the convergence speed); termination entropy value ; Generate a perturbation term through a standard normal distribution random number, calculate the new weight vector in combination with the current entropy flow intensity, and perform normalization processing to ensure that the sum of the weights is 1. Among them, the perturbation formula is:

[0035] where, w' is the new weight, w is the current weight, β is the perturbation coefficient, S k is the current entropy flow intensity, and randn() is a function that generates standard normal distribution random numbers; Calculate the entropy difference between the current state and the perturbed state, and obtain different solutions according to the entropy difference result. It should be noted that in the steps of obtaining different solutions, the calculation formula of the entropy difference is:

[0036] where, ΔE is the entropy difference, E(w') is the entropy flow energy of the perturbed state, and E(w) is the entropy flow energy of the current state; When , accept the new solution, that is, w'; When At this time, with probability , accept the new solution to avoid falling into a local optimum. Here, the new solution is also w', but at this time w' is lower than the current weight w, that is, w' is a low-precision solution; Control the dynamic entropy decay process to satisfy the exponential decay characteristic to ensure convergence within a finite number of steps. Specifically, update the entropy flow intensity after each iteration: Gradually reduce the randomness of the system to enhance convergence. It can be proved by the steady-state analysis of the Markov chain that when , the algorithm converges to the global optimal solution with probability 1; When the entropy flow intensity is less than the termination entropy value, that is , output the optimal weight vector.

[0037] Regularly evaluate the analysis results, compare them with the actual results, and verify the accuracy and reliability of the method of the present invention. Through the ROC curve analysis, the AUC value of the EFW algorithm on the test set is 0.983, which is significantly higher than 0.921 of the traditional method, proving its advantage in high-risk sample identification.

[0038] In summary, the data processing method based on cataract detection in the above embodiments of the present invention obtains historical data of different dimensions of the user and trains the AI decision model according to the historical data of different dimensions; obtains the data to be analyzed, inputs the data to be analyzed into the trained AI decision model, and outputs the analysis result. Among them, in the process of training the AI decision model, the entropy flow roaming algorithm is used to continuously adjust the weights of data of different dimensions to optimize the performance of the model. Specifically, the entropy flow roaming algorithm optimizes the weight allocation through the entropy flow potential function and the dynamic balance mechanism, can adaptively adjust the search range, is more suitable for the complex characteristics of medical data, and finally improves the accuracy and reliability of the data processing result based on cataract detection.

[0039] Embodiment 2 Please refer to Figure 2 , Figure 2 which is a structural block diagram of a data processing system based on cataract detection provided by Embodiment 2 of the present invention. The data processing system 200 based on cataract detection is used to implement the above embodiments and preferred implementation manners, and those that have been described will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that can achieve a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0040] Specifically, the data processing system 200 based on cataract detection includes: a training module 21 and an input module 22, where: A training module 21, configured to obtain historical data of a user in different dimensions, and train an AI decision-making model according to the historical data in different dimensions; An input module 22, configured to obtain data to be analyzed, input the data to be analyzed into the trained AI decision-making model, and output an analysis result; Wherein, during the training of the AI decision-making model, an entropy flow roaming algorithm is used to continuously adjust the weights of data in different dimensions to optimize the performance of the model. Specifically, the entropy flow roaming algorithm optimizes the weight allocation through an entropy flow potential function and a dynamic balance mechanism.

[0041] Further, in some optional embodiments of the present invention, the training module 21 includes: A preprocessing unit, configured to preprocess historical data in different dimensions, and perform few-sample correlation analysis on the preprocessed data to obtain a correlation analysis result. The calculation formula for the few-sample correlation analysis is:

[0042] Where d i is the rank difference between the data type and the analysis result in the i-th sample, n is the number of samples, and r s is the correlation analysis result; An allocation unit, configured to allocate initial weights to data in different dimensions according to the correlation analysis result. The initial weights are expressed as:

[0043] Where is the initial weight of the j-th type of data, r sj is the Spearman rank correlation coefficient between the j-th type of data and the cataract analysis result, r sk is the Spearman rank correlation coefficient between the k-th type of data and the cataract analysis result, and m is the total number of data types; An output unit, configured to perform a global optimal search in the solution space according to the initial weights through an adaptive switching mechanism that simulates the transition from a high-entropy state to a low-entropy state, and output an optimal weight vector of data in different dimensions.

[0044] Further, in some optional embodiments of the present invention, the output unit includes: A setting subunit, configured to define an entropy flow potential function and set an initial entropy flow intensity; A perturbation term generation subunit, configured to generate a perturbation term through a standard normal distribution random number, calculate a new weight vector in combination with the current entropy flow intensity, and perform a normalization process to ensure that the sum of the weights is 1; A calculation subunit, configured to calculate the entropy difference between the current state and the perturbed state, and obtain different solutions according to the entropy difference result; A control subunit for controlling the dynamic entropy decay process to satisfy the exponential decay characteristic, so as to ensure convergence within a finite number of steps; An output subunit for outputting an optimal weight vector when the entropy flow intensity is less than the termination entropy value.

[0045] Embodiment III On the other hand, the present invention also proposes an electronic device. Please refer to Figure 3 , which shows the electronic device in Embodiment III of the present invention, including a memory 20, a processor 10, and a computer program 30 stored on the memory and executable on the processor. When the processor 10 executes the computer program 30, it implements the data processing method based on cataract detection as described above.

[0046] Among them, in some embodiments, the processor 10 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips, and is used to run the program code stored in the memory 20 or process data, such as executing an access restriction program, etc.

[0047] Among them, the memory 20 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, a hard disk, a multimedia card, a card-type memory (for example, an SD or DX memory, etc.), a magnetic memory, a magnetic disk, an optical disk, etc. The memory 20 may be an internal storage unit of the electronic device in some embodiments, such as the hard disk of the electronic device. The memory 20 may also be an external storage device of the electronic device in other embodiments, such as a plug-in hard disk equipped on the electronic device, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. Further, the memory 20 may also include both an internal storage unit and an external storage device of the electronic device. The memory 20 can not only be used to store application software and various types of data of the electronic device, but also be used to temporarily store data that has been output or will be output.

[0048] It should be noted that Figure 3 the structure shown does not constitute a limitation on the electronic device. In other embodiments, the electronic device may include fewer or more components than shown, or combine certain components, or have a different component layout.

[0049] The embodiment of the present invention also proposes a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the data processing method based on cataract detection as described above.

[0050] Those skilled in the art will understand that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a definite sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or used in combination with these instruction execution systems, apparatus, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.

[0051] More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection part (electronic device) having one or more wirings, a portable computer disk cartridge (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or other suitable processing as necessary, and then stored in a computer memory.

[0052] It should be understood that various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application-specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0053] In the description of this specification, the descriptions referring to terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples", etc., mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0054] The above embodiments merely represent several implementation manners of the present invention. The description thereof is relatively specific and detailed, but it should not be construed as a limitation on the scope of the patent for the present invention. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all fall within the protection scope of the present invention. Therefore, the protection scope of the patent for the present invention shall be subject to the appended claims.

Claims

1. A data processing method based on cataract detection, characterized in that: The method comprises: Obtaining historical data of different dimensions of the user for cataract detection, and training the AI ​​decision model according to the historical data of different dimensions, wherein the historical data at least includes basic user data and eye data related to cataract detection; Obtaining data to be analyzed, inputting the data to be analyzed into a trained AI decision model, and outputting analysis results; In the process of training the AI ​​decision-making model, the entropy flow roaming algorithm is used to continuously adjust the weights of data of different dimensions to optimize the performance of the model. Specifically, the historical data of different dimensions are preprocessed, and the preprocessed data is subjected to a small sample size correlation analysis to obtain the correlation analysis results; According to the correlation analysis results, initial weights are assigned to data of different dimensions; According to the initial weights, by simulating an adaptive switching mechanism from a high entropy state to a low entropy state, a global optimal search is implemented in the solution space, and optimal weight vectors of data of different dimensions are output.

2. The data processing method based on cataract detection according to claim 1, characterized in that: In the step of performing a small sample size correlation analysis on the preprocessed data to obtain a correlation analysis result, the calculation formula for the small sample size correlation analysis is: Among them, d i is the grade difference between the data type and the analysis result in the i-th sample, n is the number of samples, r s The results of correlation analysis.

3. The data processing method based on cataract detection according to claim 2, characterized in that: In the step of assigning initial weights to data of different dimensions according to the correlation analysis results, the initial weights are expressed as: in, is the initial weight of the j-th class of data, r sj is the Spearman rank correlation coefficient between the j-th type of data and the cataract analysis results, r sk is the Spearman rank correlation coefficient between the k-th type of data and the cataract analysis results, and m is the total number of data types.

4. The data processing method based on cataract detection according to claim 3, characterized in that: The step of implementing a global optimal search in the solution space and outputting optimal weight vectors of data of different dimensions by simulating an adaptive switching mechanism from a high entropy state to a low entropy state according to the initial weights comprises: Define the entropy flow potential function and set the initial entropy flow intensity; Generate the disturbance term through the standard normal distribution random number, calculate the new weight vector based on the current entropy flow intensity, and perform normalization to ensure that the sum of the weights is 1; Calculate the entropy difference between the current state and the disturbed state, and obtain different solutions based on the entropy difference results; Control the dynamic entropy decay process to satisfy the exponential decay characteristic to ensure convergence within a finite number of steps; When the entropy flow intensity is less than the termination entropy value, the optimal weight vector is output.

5. The data processing method based on cataract detection according to claim 4, characterized in that: In the step of generating a disturbance term by using a standard normal distribution random number and calculating a new weight vector in combination with the current entropy flow intensity, the disturbance formula is: Among them, w' is the new weight, w is the current weight, β is the perturbation coefficient, S k is the current entropy flow intensity, and randn() is a function for generating standard normal distribution random numbers.

6. The data processing method based on cataract detection according to claim 5, characterized in that: In the step of calculating the entropy difference between the current state and the disturbance state, and obtaining different solutions according to the entropy difference result, the calculation formula of the entropy difference is: Among them, ΔE is the entropy difference, E(w') is the entropy flow energy of the disturbed state, and E(w) is the entropy flow energy of the current state; when When accepting new interpretations; when When, with probability , accept the new interpretation.

7. A data processing system based on cataract detection, characterized in that: Used to implement the data processing method based on cataract detection according to any one of claims 1 to 6, the system comprises: The training module is used to obtain historical data of users in different dimensions and train the AI ​​decision model based on the historical data in different dimensions; An input module is used to obtain data to be analyzed, input the data to be analyzed into a trained AI decision model, and output analysis results; In the process of training the AI ​​decision-making model, the entropy flow roaming algorithm is used to continuously adjust the weights of data of different dimensions to optimize the performance of the model. Specifically, the historical data of different dimensions are preprocessed, and the preprocessed data is subjected to a small sample size correlation analysis to obtain the correlation analysis results; According to the correlation analysis results, initial weights are assigned to data of different dimensions; According to the initial weights, by simulating an adaptive switching mechanism from a high entropy state to a low entropy state, a global optimal search is implemented in the solution space, and optimal weight vectors of data of different dimensions are output.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the data processing method based on cataract detection as described in any one of claims 1 to 6 is implemented.

9. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method for data processing based on cataract detection as claimed in any one of claims 1 to 6 is implemented.

Images