Glaucoma differentiation detection method and system based on multimodal fusion

Abstract

The present invention discloses a glaucoma differentiation detection method and system based on multimodal fusion. The method relates to the technical field of glaucoma differentiation detection, and includes the following steps: collecting glaucoma differentiation detection data; analyzing glaucoma differentiation detection data; and comprehensively analyzing and optimizing the method for adjusting the quality accuracy of glaucoma differentiation detection data. The present invention collects and processes glaucoma differentiation detection data, analyzes the glaucoma differentiation detection data, obtains a glaucoma differentiation detection data integrity assessment index and a glaucoma differentiation detection data image quality assessment index, comprehensively analyzes and optimizes the method for adjusting the quality accuracy of glaucoma differentiation detection data, thereby improving the accuracy of the glaucoma differentiation detection method and system based on multimodal fusion, and solving the problem of insufficient accuracy of the glaucoma differentiation detection method and system based on multimodal fusion in the prior art.

Description

Technical Field

[0001] The present invention relates to the technical field of glaucoma differentiation detection, and in particular to a glaucoma differentiation detection method and system based on multimodal fusion. Background Art

[0002] Glaucoma is the leading cause of irreversible blindness worldwide, characterized by progressive loss of retinal ganglion cells, structural changes in the optic nerve head, and corresponding visual function defects. Early detection and preventive treatment are crucial to reducing vision loss caused by glaucoma. Traditional glaucoma diagnosis mainly relies on subjective analysis by clinicians, such as observing fundus images and OCT images, which requires a large number of experienced experts and is time-consuming and labor-intensive. Single-modality data analysis is often unreliable. The diagnostic process of glaucoma requires a joint functional and structural analysis. With the optimization of algorithms, upgrades to cloud services, and the accumulation of data volume, the successful application of deep learning-based artificial intelligence technology in image, video, voice and other fields has attracted the attention of ophthalmologists. Multimodal fusion technology is expected to improve the accuracy and efficiency of glaucoma diagnosis, facilitate early detection and treatment, and thus reduce the risk of visual impairment and blindness.

[0003] The existing glaucoma differentiation detection system based on multimodal fusion uses multimodal imaging technology to observe the structural changes of the retina and optic nerve head; uses deep learning algorithms, especially Transformer-based models, through deep learning and artificial intelligence to extract and fuse features of multimodal imaging data; uses a glaucoma grading method based on spatial relationship multimodal feature fusion, uses a 3D Transformer model to extract features from the OCT volume, and combines them with the features of the fundus image for fusion.

[0004] For example, the invention patent with announcement number: CN106308738B announces a device and method for glaucoma detection, including: a table display mechanism and a sub-table detection mechanism, and a mounting connector connecting the table display mechanism and the sub-table detection mechanism; the table display mechanism includes a front shell, an image display, a camera monitor, a medical display electrically connected to the image display, and a camera monitor electrically connected to the camera monitor, which are arranged on the front shell; the image display and the camera monitor are arranged on the front shell via a first mounting bracket, and the medical display and the camera monitor are arranged on the front shell via a second mounting bracket.

[0005] For example, the invention patent application with publication number: CN116236151A discloses a wearable detection method and device for glaucoma disease risk, including: S1 obtains the user's movement posture, photoelectric volume pulse wave signal and light information of the user wearing the bracelet; S2 calculates the heart rate, blood oxygen, and blood pressure information through the photoelectric volume pulse wave signal, and transmits the user's movement posture, heart rate, blood oxygen, blood pressure and light data information to the glaucoma risk identification model; S3 uses the glaucoma risk identification model to integrate and analyze the user's movement posture, heart rate, blood oxygen, blood pressure and light data information to obtain the probability that the user suffers from glaucoma; so that patients do not need to draw blood or go to the hospital for intraocular pressure examination, and can realize glaucoma disease risk detection in daily life, thereby effectively helping the screening of glaucoma disease.

[0006] However, in the process of implementing the technical solution of the invention in the embodiments of the present application, the present application found that the above technology has at least the following technical problems:

[0007] This method has the problem of insufficient accuracy of the quality of glaucoma differentiation detection data of multimodal fusion. During the storage and transmission of glaucoma differentiation detection data, the glaucoma differentiation detection data is incomplete due to deviations when compressing images or performing format conversion. This leads to increased noise and artifacts in glaucoma image detection, and the quality of glaucoma images of multimodal fusion is inaccurate. There is a problem of insufficient accuracy of the glaucoma differentiation detection method and system based on multimodal fusion. Summary of the invention

[0008] The embodiments of the present application solve the problem of insufficient accuracy of the glaucoma differentiation detection method and system based on multimodal fusion in the prior art by providing a glaucoma differentiation detection method and system based on multimodal fusion, and improve the accuracy of the glaucoma differentiation detection method and system based on multimodal fusion.

[0009] The embodiment of the present application provides a glaucoma differentiation detection method based on multimodal fusion, comprising the following steps: collecting and processing glaucoma differentiation detection data; analyzing the glaucoma differentiation detection data to obtain a glaucoma differentiation detection data integrity evaluation index and a glaucoma differentiation detection data image quality evaluation index; comprehensively analyzing to obtain a glaucoma differentiation detection data quality accuracy evaluation index; comparing and analyzing the glaucoma differentiation detection data integrity evaluation index with a first threshold of the glaucoma differentiation detection data integrity evaluation index, and optimizing and adjusting the glaucoma differentiation detection data integrity method; comparing and analyzing the glaucoma differentiation detection data image quality evaluation index with a second threshold of the glaucoma differentiation detection data image quality evaluation index, and optimizing and adjusting the glaucoma differentiation detection data image quality method; comparing and analyzing the glaucoma differentiation detection data quality accuracy evaluation index with a comprehensive threshold of the glaucoma differentiation detection data quality accuracy evaluation index, and optimizing and adjusting the glaucoma differentiation detection data quality accuracy method.

[0010] Furthermore, the specific steps of collecting and processing glaucoma differentiation detection data are: collecting glaucoma differentiation detection raw data through sensor equipment; cleaning and denoising the glaucoma differentiation detection raw data to obtain glaucoma differentiation detection data, and the glaucoma differentiation detection data includes glaucoma differentiation detection data integrity data and glaucoma differentiation detection data image quality data.

[0011] Furthermore, the specific steps of obtaining the glaucoma differentiation detection data integrity evaluation index are: obtaining the glaucoma differentiation detection data entropy value of the preset glaucoma differentiation detection data integrity detection point through image processing software; obtaining the run length of the preset glaucoma differentiation detection data integrity detection point through image compression software; obtaining the maximum value of the peak signal intensity of the preset glaucoma differentiation detection data integrity detection point through image processing software; obtaining the minimum value of the peak signal intensity of the preset glaucoma differentiation detection data integrity detection point through image processing software; obtaining the image noise of the preset glaucoma differentiation detection data integrity detection point through a noise analysis tool; obtaining the image artifact intensity of the preset glaucoma differentiation detection data integrity detection point through image processing software; obtaining the ambient light angle of the preset glaucoma differentiation detection data integrity detection point through a light measurement device; the glaucoma differentiation detection data integrity data includes the glaucoma differentiation detection data entropy value, run length, maximum value of peak signal intensity, minimum value of peak signal intensity, image noise, image artifact intensity and ambient light angle; and obtaining the glaucoma differentiation detection data integrity evaluation index according to the glaucoma differentiation detection data integrity data analysis.

[0012] Furthermore, the specific steps for obtaining the image quality evaluation index of the glaucoma differentiation detection data are: obtaining the image compression ratio of the preset glaucoma differentiation detection data image quality detection point through image compression software; obtaining the run length of the preset glaucoma differentiation detection data image quality detection point through image compression software; obtaining the image contrast of the preset glaucoma differentiation detection data image quality detection point through image processing software; obtaining the image brightness of the preset glaucoma differentiation detection data image quality detection point through image processing software; obtaining the image noise of the preset glaucoma differentiation detection data image quality detection point through image processing software; obtaining the image artifact intensity of the preset glaucoma differentiation detection data image quality detection point through image processing software; the image quality data of the glaucoma differentiation detection data includes image compression ratio, run length, image contrast, image brightness, image noise, and image artifact intensity; and obtaining the image quality evaluation index of the glaucoma differentiation detection data according to the image quality data analysis of the glaucoma differentiation detection data.

[0013] Furthermore, the specific steps of the comprehensive analysis to obtain the glaucoma differentiation detection data quality accuracy evaluation index are: obtaining the image resolution of the preset glaucoma differentiation detection data quality accuracy detection point through image processing software; and obtaining the glaucoma differentiation detection data quality accuracy evaluation index through comprehensive analysis of the glaucoma differentiation detection data integrity evaluation index, the glaucoma differentiation detection data image quality evaluation index and the image resolution.

[0014] Furthermore, the specific steps of optimizing and adjusting the glaucoma differentiation detection data integrity method are as follows: if the glaucoma differentiation detection data integrity assessment index is greater than or equal to the first threshold value of the glaucoma differentiation detection data integrity assessment index, there is no need to optimize and adjust the glaucoma differentiation detection data integrity method; if the glaucoma differentiation detection data integrity assessment index is lower than the first threshold value of the glaucoma differentiation detection data integrity assessment index, then the difference between the glaucoma differentiation detection data integrity assessment index and the first threshold value of the glaucoma differentiation detection data integrity assessment index is used to match the corresponding adjustment plan in the glaucoma differentiation detection database.

[0015] Furthermore, the specific steps of the method for optimizing and adjusting the image quality of glaucoma differentiation detection data are as follows: if the image quality assessment index of the glaucoma differentiation detection data is lower than or equal to the second threshold value of the image quality assessment index of the glaucoma differentiation detection data, there is no need to optimize and adjust the image quality method of the glaucoma differentiation detection data; if the image quality assessment index of the glaucoma differentiation detection data is greater than the second threshold value of the image quality assessment index of the glaucoma differentiation detection data, then the difference between the image quality assessment index of the glaucoma differentiation detection data and the second threshold value of the image quality assessment index of the glaucoma differentiation detection data is used to match the corresponding adjustment plan in the glaucoma differentiation detection database.

[0016] Furthermore, the specific steps of the method for optimizing and adjusting the quality accuracy of glaucoma differentiation detection data are: extracting the comprehensive threshold of the glaucoma differentiation detection data quality accuracy assessment index in the glaucoma differentiation detection database, and comparing the glaucoma differentiation detection data quality accuracy assessment index with the comprehensive threshold of the glaucoma differentiation detection data quality accuracy assessment index; if the glaucoma differentiation detection data quality accuracy assessment index is greater than or equal to the comprehensive threshold of the glaucoma differentiation detection data quality accuracy assessment index, there is no need to optimize and adjust the quality accuracy of the glaucoma differentiation detection data.

[0017] Furthermore, the method for optimizing and adjusting the quality accuracy of glaucoma differentiation detection data also includes: if the glaucoma differentiation detection data quality accuracy assessment index is lower than the comprehensive threshold of the glaucoma differentiation detection data quality accuracy assessment index, then matching the corresponding adjustment plan in the glaucoma differentiation detection database through the difference between the glaucoma differentiation detection data quality accuracy assessment index and the comprehensive threshold of the glaucoma differentiation detection data quality accuracy assessment index.

[0018] The embodiment of the present application provides a glaucoma differentiation detection system based on multimodal fusion, including a glaucoma differentiation detection data collection module, a glaucoma differentiation detection data analysis module, a comprehensive analysis module and a glaucoma differentiation detection data quality accuracy method module: a glaucoma differentiation detection data collection module: used to collect and process glaucoma differentiation detection data; a glaucoma differentiation detection data analysis module: used to analyze the glaucoma differentiation detection data to obtain a glaucoma differentiation detection data integrity evaluation index and a glaucoma differentiation detection data image quality evaluation index; a comprehensive analysis module: used to comprehensively analyze and obtain a glaucoma differentiation detection data quality accuracy evaluation index ; Module for optimizing and adjusting the quality accuracy method of glaucoma differentiation detection data: used to compare and analyze the glaucoma differentiation detection data integrity assessment index with the first threshold of the glaucoma differentiation detection data integrity assessment index, and optimize and adjust the glaucoma differentiation detection data integrity method; compare and analyze the glaucoma differentiation detection data image quality assessment index with the second threshold of the glaucoma differentiation detection data image quality assessment index, and optimize and adjust the glaucoma differentiation detection data image quality method; compare and analyze the glaucoma differentiation detection data quality accuracy assessment index with the comprehensive threshold of the glaucoma differentiation detection data quality accuracy assessment index, and optimize and adjust the glaucoma differentiation detection data quality accuracy method.

[0019] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0020] 1. By collecting and processing glaucoma differentiation detection data, analyzing the glaucoma differentiation detection data, obtaining the glaucoma differentiation detection data integrity evaluation index and the glaucoma differentiation detection data image quality evaluation index, comprehensively analyzing and optimizing the method for adjusting the glaucoma differentiation detection data quality accuracy, and applying the filtering operation multiple times through iterative filters to gradually improve the image quality. Each iteration can adjust the filtering parameters according to the previous results to improve the accuracy of glaucoma differentiation detection, thereby solving the problem of insufficient accuracy of glaucoma differentiation detection methods and systems based on multimodal fusion in the prior art;

[0021] 2. By analyzing the glaucoma differentiation detection data, the glaucoma differentiation detection data integrity evaluation index and the glaucoma differentiation detection data image quality evaluation index are obtained. By using edge-preserving filters to enhance the key features in the glaucoma differentiation detection image and reduce image noise, by using multi-scale analysis to remove noise and retain important image features, the quality of the glaucoma differentiation detection image is improved;

[0022] 3. Through comprehensive analysis and optimization of the method to adjust the data quality accuracy of glaucoma differentiation detection, optimize the filter parameters, retain the edges and details of the image, and gradually improve the image quality by applying the filtering operation multiple times through iterative filters, the image quality accuracy of glaucoma differentiation detection is improved, thereby improving the glaucoma differentiation detection method based on multimodal fusion and the system reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A flow chart of a glaucoma differentiation and detection method based on multimodal fusion provided in an embodiment of the present application;

[0024] Figure 2 A schematic diagram of an index function for evaluating the quality accuracy of glaucoma differentiation detection data provided in an embodiment of the present application;

[0025] Figure 3 A schematic diagram of the structure of a glaucoma differentiation and detection system based on multimodal fusion provided in an embodiment of the present application. DETAILED DESCRIPTION

[0026] The embodiments of the present application solve the problem of insufficient accuracy of the glaucoma differentiation detection method and system based on multimodal fusion in the prior art by providing a glaucoma differentiation detection method and system based on multimodal fusion. By collecting and processing glaucoma differentiation detection data, analyzing the glaucoma differentiation detection data, obtaining a glaucoma differentiation detection data integrity assessment index and a glaucoma differentiation detection data image quality assessment index, comprehensively analyzing and optimizing the method for adjusting the quality accuracy of glaucoma differentiation detection data, the accuracy of the glaucoma differentiation detection method and system based on multimodal fusion is improved.

[0027] The technical solution in the embodiment of the present application is to solve the above-mentioned problem that the glaucoma differentiation detection method and system based on multimodal fusion are insufficient in accuracy. The overall idea is as follows:

[0028] By collecting and processing glaucoma differentiation detection data, analyzing the glaucoma differentiation detection data, obtaining the glaucoma differentiation detection data integrity assessment index and the glaucoma differentiation detection data image quality assessment index, comprehensively analyzing and optimizing the glaucoma differentiation detection data quality accuracy method, the accuracy of the glaucoma differentiation detection method and system based on multimodal fusion is improved.

[0029] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0030] like Figure 1 As shown, it is a flow chart of the glaucoma differentiation detection method based on multimodal fusion provided in an embodiment of the present application, the method is applied to a glaucoma differentiation detection system based on multimodal fusion, and the method includes the following steps: collecting and processing glaucoma differentiation detection data; analyzing the glaucoma differentiation detection data to obtain a glaucoma differentiation detection data integrity evaluation index and a glaucoma differentiation detection data image quality evaluation index; comprehensively analyzing to obtain a glaucoma differentiation detection data quality accuracy evaluation index; comparing and analyzing the glaucoma differentiation detection data integrity evaluation index with a first threshold of the glaucoma differentiation detection data integrity evaluation index, and optimizing and adjusting the glaucoma differentiation detection data integrity method; comparing and analyzing the glaucoma differentiation detection data image quality evaluation index with a second threshold of the glaucoma differentiation detection data image quality evaluation index, and optimizing and adjusting the glaucoma differentiation detection data image quality method; comparing and analyzing the glaucoma differentiation detection data quality accuracy evaluation index with a comprehensive threshold of the glaucoma differentiation detection data quality accuracy evaluation index, and optimizing and adjusting the glaucoma differentiation detection data quality accuracy method.

[0031] In this embodiment, the multimodal data includes text and image data, wherein if there is a missing value in the text data, the missing value can be obtained by the following method, for example, the linear correlation between the clinical data data is compared by using the Person coefficient, and the parameters with greater correlation with the missing value are selected from the linear filling model for filling, so as to obtain the estimated value of the missing value in the clinical data data. Specifically, when filling the data, thousands of data can be filled in about 3 minutes. The advantage of this step is that the present invention can fill the data in a relatively low power consumption, low overhead and short time. The filled clinical data data is placed in the corresponding feature extraction network to obtain the feature G, and the RNFL deviation map, RNFL thickness map and optic nerve cross-sectional map in the OCT equipment instrument are respectively extracted on the feature extraction network to obtain the features T, A and E. Specifically, this step extracts features from data of different modalities, and can effectively extract data of different modalities. The clinical data data feature G, the RNFL deviation map feature T, the RNFL thickness map feature A and the optic nerve cross-sectional map feature E are feature fused through the corresponding hyperparameters. Specifically, this step performs adaptive feature fusion by setting corresponding hyperparameters for the data of each modality. The trained multimodal fusion network is used for feature inversion. Pathological analysis of primary open-angle glaucoma and primary angle-closure glaucoma is performed by means of heat maps. This step uses the trained multimodal fusion network to visualize the features of primary open-angle glaucoma and primary angle-closure glaucoma by gradient descent.

[0032] Furthermore, the specific steps of collecting and processing glaucoma differentiation detection data are: collecting glaucoma differentiation detection raw data through sensor equipment; cleaning and denoising the glaucoma differentiation detection raw data to obtain glaucoma differentiation detection data, wherein the glaucoma differentiation detection data includes glaucoma differentiation detection data integrity data and glaucoma differentiation detection data image quality data.

[0033] Furthermore, the specific steps for obtaining the glaucoma differentiation detection data integrity evaluation index are as follows: obtaining the glaucoma differentiation detection data entropy value of the preset glaucoma differentiation detection data integrity detection point through image processing software; obtaining the run length of the preset glaucoma differentiation detection data integrity detection point through image compression software; obtaining the maximum value of the peak signal intensity of the preset glaucoma differentiation detection data integrity detection point through image processing software; obtaining the minimum value of the peak signal intensity of the preset glaucoma differentiation detection data integrity detection point through image processing software; obtaining the image noise of the preset glaucoma differentiation detection data integrity detection point through a noise analysis tool; obtaining the image artifact intensity of the preset glaucoma differentiation detection data integrity detection point through image processing software; obtaining the ambient light angle of the preset glaucoma differentiation detection data integrity detection point through a light measurement device; the glaucoma differentiation detection data integrity data includes the glaucoma differentiation detection data entropy value, run length, maximum value of peak signal intensity, minimum value of peak signal intensity, image noise, image artifact intensity and ambient light angle; and obtaining the glaucoma differentiation detection data integrity evaluation index according to the glaucoma differentiation detection data integrity data analysis.

[0034] In this embodiment, the specific method for analyzing and obtaining the glaucoma differentiation detection data integrity evaluation index is as follows:

[0035] ;

[0036] ; ;

[0037] ; ;

[0038] ; ;

[0039] The preset glaucoma differentiation test data integrity test points are numbered in sequence. Indicates The number of the glaucoma differentiation detection data integrity detection point under the glaucoma differentiation detection data integrity detection segment, , The total number of test points indicating the data integrity of the glaucoma differentiation test.

[0040] Divide the preset glaucoma differentiation detection data integrity time into glaucoma differentiation detection data integrity detection segments of equal length, Indicates the number of the glaucoma differentiation test data integrity test segment, , The total number of test segments indicating the integrity of the glaucoma differentiation test data;

[0041] Indicates The glaucoma differentiation test data integrity assessment index for each glaucoma differentiation test data integrity test point;

[0042] Indicates Image noise correction coefficient at each glaucoma differentiation test data integrity test point;

[0043] Indicates Image artifact intensity correction coefficient at each glaucoma differentiation test data integrity test point;

[0044] Indicates The ambient light angle correction factor at each glaucoma differentiation test data integrity test point;

[0045] Indicates The entropy value of glaucoma differentiation detection data under each glaucoma differentiation detection data integrity detection point;

[0046] represents the entropy value threshold of glaucoma differentiation detection data, which is a preset entropy value threshold of glaucoma differentiation detection data obtained from the glaucoma differentiation detection database, and may be an average value of the entropy value of glaucoma differentiation detection data at a preset glaucoma differentiation detection data integrity detection point from a historical database;

[0047] Indicates the standard value of the stroke length, which is a preset standard value of the stroke length obtained from the glaucoma differentiation detection database, and can be the average value of the stroke length under the preset glaucoma differentiation detection data integrity detection point from the historical database. During the transmission of the glaucoma differentiation detection data, the setting of the stroke length will affect the integrity of the glaucoma differentiation detection data. If the deviation between the stroke length and the standard value of the stroke length is large, errors will occur in the transmission of the glaucoma differentiation detection data, thereby affecting the data integrity evaluation index;

[0048] Indicates The run length under the glaucoma differentiation test data integrity test point;

[0049] Indicates The frequency domain difference coefficient under the glaucoma differentiation test data integrity test segment;

[0050] Indicates Image noise at each glaucoma differentiation test data integrity test point;

[0051] represents the image noise standard value, which is the preset image noise standard value obtained from the glaucoma differentiation detection database, and can be the image noise average value at the preset glaucoma differentiation detection data integrity detection point from the historical database;

[0052] Indicates The intensity of image artifacts at each glaucoma differentiation test data integrity test point;

[0053] represents the standard value of image artifact intensity, which is a preset standard value of image artifact intensity obtained from the glaucoma differentiation detection database, and can be the average value of image artifact intensity at the preset glaucoma differentiation detection data integrity detection point from the historical database;

[0054] Indicates The ambient light angle at each glaucoma differentiation test data integrity test point;

[0055] represents the standard value of ambient light angle, which is the preset standard value of ambient light angle obtained from the glaucoma differentiation detection database, and can be the average value of ambient light angle under the preset glaucoma differentiation detection data integrity detection point from the historical database;

[0056] Indicates The maximum value of the peak signal intensity under the glaucoma differentiation test data integrity test segment;

[0057] Indicates The minimum value of peak signal intensity under the glaucoma differentiation test data integrity test segment;

[0058] represents the standard value of the error peak signal strength, which is the preset standard value of the error peak signal strength obtained from the glaucoma differentiation detection database, and can be the average value of the error peak signal strength under the preset glaucoma differentiation detection data integrity detection segment from the historical database, and e represents a natural constant;

[0059] is a weight factor influencing the entropy value of the preset glaucoma differentiation detection data obtained from the glaucoma differentiation detection database;

[0060] A preset run length impact weight factor obtained from a glaucoma differentiation detection database;

[0061] The weighting factor is a preset frequency domain difference coefficient obtained from the glaucoma differentiation detection database;

[0062] is a preset image noise correction factor obtained from the glaucoma differentiation detection database;

[0063] is a preset image artifact intensity correction factor obtained from the glaucoma differentiation detection database;

[0064] is the preset ambient light angle correction factor obtained from the glaucoma differentiation detection database.

[0065] The preset glaucoma differentiation detection data entropy value influence weight factor, the preset run length influence weight factor and the preset frequency domain difference coefficient influence weight factor respectively represent the numerical value of the influence of the glaucoma differentiation detection data entropy value, run length and frequency domain difference coefficient on the glaucoma differentiation detection data integrity assessment index, and represent the proportion of the glaucoma differentiation detection data entropy value, run length and frequency domain difference coefficient in the influence on the glaucoma differentiation detection data integrity assessment index.

[0066] The preset glaucoma differentiation detection data entropy value influencing weight factor, the preset stroke length influencing weight factor and the preset frequency domain difference coefficient influencing weight factor are obtained through mapping relationships. For example, a mapping set of glaucoma differentiation detection data entropy value, stroke length and frequency domain difference coefficient and their corresponding weights are established respectively through the relationship between the glaucoma differentiation detection data entropy value, stroke length and frequency domain difference coefficient in the historical data and the data update frequency. The corresponding preset glaucoma differentiation detection data entropy value influencing weight factor, preset stroke length influencing weight factor and preset frequency domain difference coefficient influencing weight factor in the mapping set are obtained by inputting the real-time glaucoma differentiation detection data entropy value, stroke length and frequency domain difference coefficient.

[0067] The preset image noise correction factor, the preset image artifact intensity correction factor and the preset ambient light angle correction factor respectively represent the numerical values ​​of the influence of image noise, image artifact intensity and ambient light angle on the glaucoma differentiation detection data integrity assessment index, and represent the proportion of image noise, image artifact intensity and ambient light angle in the influence of the glaucoma differentiation detection data integrity assessment index.

[0068] The preset image noise correction factor, the preset image artifact intensity correction factor and the preset ambient light angle correction factor are obtained through mapping relationships. For example, a mapping set of image noise, image artifact intensity and ambient light angle and their corresponding weights is established respectively through the relationship between image noise, image artifact intensity and ambient light angle in historical data, and the preset image noise correction factor, preset image artifact intensity correction factor and preset ambient light angle correction factor corresponding to the mapping set are obtained by inputting real-time image noise, image artifact intensity and ambient light angle.

[0069] The larger the entropy value of glaucoma differentiation detection data, the more random noise there is in the image, and the larger the image noise; the maximum and minimum values ​​of the peak signal intensity reflect the brightness range in the image, which is related to the image contrast. The higher the image contrast, the larger the square of the difference between the maximum value of the peak signal intensity and the maximum value of the peak signal intensity. The image noise and image artifact intensity are factors that affect the image quality, and the image noise increases; the image artifact intensity is more obvious, and the presence of image artifacts will increase the noise level of the image, and the larger the image artifact intensity; the ambient illumination angle will affect the brightness and contrast of the image, thereby affecting the image quality; the smaller the deviation of the ambient illumination angle from the ambient standard value, the more visible the details of the image can be; the relationship between the maximum and minimum values ​​of the peak signal intensity and the image noise depends on the brightness and contrast of the image. The larger the image noise, the more it will affect the accuracy of the peak signal intensity; the ambient illumination angle will affect the noise level of the image. The larger the deviation of the ambient illumination angle from the ambient illumination angle standard value, the reflection and shadow in the image will increase the noise, and the larger the image noise.

[0070] The entropy value of glaucoma differentiation detection data reflects the complexity and information content of glaucoma differentiation detection images, and is negatively correlated with the glaucoma differentiation detection data integrity assessment index. The larger the absolute value of the difference between the entropy value of glaucoma differentiation detection data and the standard value of the entropy value of glaucoma differentiation detection data, the smaller the glaucoma differentiation detection data integrity assessment index; the run length reduces the amount of data by the length of consecutive identical pixel values, and is negatively correlated with the glaucoma differentiation detection data integrity assessment index. The larger the absolute value of the difference between the run length and the standard value of the run length, the smaller the glaucoma differentiation detection data integrity assessment index; the maximum and minimum values ​​of the peak signal intensity reflect the brightness range in the image, which is related to the image contrast. Images with high contrast usually have a larger peak signal intensity difference, and are negatively correlated with the glaucoma differentiation detection data integrity assessment index. The larger the square of the difference between the maximum peak signal intensity and the minimum peak signal intensity, the smaller the glaucoma differentiation detection data integrity assessment index. The smaller the image noise is, the larger the glaucoma differentiation detection data integrity assessment index is; there is a negative correlation between the square of the difference between image noise and the image noise standard value and the glaucoma differentiation detection data integrity assessment index. The larger the square of the difference between image noise and the image noise standard value, the more it will mask the details of the image, and the smaller the glaucoma differentiation detection data integrity assessment index is; there is a negative correlation between the square of the difference between image artifact intensity and the image artifact intensity standard value and the glaucoma differentiation detection data integrity assessment index. The larger the square of the difference between image artifact intensity and the image artifact intensity standard value, the more it interferes with the correct interpretation of the image content, and the smaller the glaucoma differentiation detection data integrity assessment index is; there is a negative correlation between the square of the difference between the ambient lighting angle and the ambient lighting angle standard value and the glaucoma differentiation detection data integrity assessment index. The larger the square of the difference between the ambient lighting angle and the ambient lighting angle standard value, the more it will cause image distortion, and the smaller the glaucoma differentiation detection data integrity assessment index is.

[0071] Furthermore, the specific steps for obtaining the image quality evaluation index of glaucoma differentiation detection data are as follows: obtaining the image compression ratio of the preset glaucoma differentiation detection data image quality detection point through image compression software; obtaining the run length of the preset glaucoma differentiation detection data image quality detection point through image compression software; obtaining the image contrast of the preset glaucoma differentiation detection data image quality detection point through image processing software; obtaining the image brightness of the preset glaucoma differentiation detection data image quality detection point through image processing software; obtaining the image noise of the preset glaucoma differentiation detection data image quality detection point through image processing software; obtaining the image artifact intensity of the preset glaucoma differentiation detection data image quality detection point through image processing software; the image quality data of the glaucoma differentiation detection data includes image compression ratio, run length, image contrast, image brightness, image noise, and image artifact intensity; and obtaining the image quality evaluation index of the glaucoma differentiation detection data according to the image quality data analysis of the glaucoma differentiation detection data.

[0072] In this embodiment, the specific method for analyzing and obtaining the image quality evaluation index of glaucoma differentiation detection data is as follows:

[0073] ;

[0074] ; ;

[0075] ; ;

[0076] ;

[0077] The preset glaucoma differentiation detection data image quality detection points are numbered in sequence. Indicates the number of the image quality detection point of the glaucoma differentiation detection data, , Indicates the total number of numbers of image quality detection points of glaucoma differentiation detection data;

[0078] Indicates The glaucoma differentiation detection data image quality assessment index of each glaucoma differentiation detection data image quality detection point;

[0079] Indicates Image compression ratio of glaucoma differentiation detection data at image quality detection points;

[0080] represents the standard value of image compression ratio, which is the preset standard value of image compression ratio obtained from the glaucoma differentiation detection database, and can be the average value of image compression ratio under the preset glaucoma differentiation detection data integrity detection point from the historical database;

[0081] Indicates The run length under the image quality detection point of glaucoma differentiation detection data;

[0082] Indicates the standard value of the stroke length, which is a preset standard value of the stroke length obtained from the glaucoma differentiation detection database, and can be the average value of the stroke length under the image quality detection point of the preset glaucoma differentiation detection data from the historical database. If the deviation between the stroke length and the standard value of the stroke length is large, the glaucoma differentiation detection data will be wrong during the transmission process, thereby affecting the image quality;

[0083] Indicates Image noise influence coefficient at each glaucoma differentiation detection data image quality detection point;

[0084] Indicates Image artifact intensity influence coefficient at each glaucoma differentiation detection data image quality detection point;

[0085] Indicates Image structure similarity coefficient of glaucoma differentiation detection data under image quality detection points;

[0086] Indicates Image noise at each glaucoma differentiation detection data image quality detection point;

[0087] represents the image noise standard value, which is a preset image noise standard value obtained from the glaucoma differentiation detection database, and can be the image noise average value at the image quality detection point of the preset glaucoma differentiation detection data from the historical database;

[0088] Indicates The image artifact intensity at each glaucoma differentiation detection data image quality detection point;

[0089] represents the standard value of image artifact intensity, which is a preset standard value of image artifact intensity obtained from the glaucoma differentiation detection database, and can be the average value of image artifact intensity at the image quality detection point of the preset glaucoma differentiation detection data from the historical database;

[0090] Indicates Image contrast at each glaucoma differentiation test data image quality test point;

[0091] Indicates Image brightness at each glaucoma differentiation test data image quality test point;

[0092] represents the image contrast standard value, which is a preset image contrast standard value obtained from the glaucoma differentiation detection database, and can be the image contrast average value at the image quality detection point of the preset glaucoma differentiation detection data from the historical database;

[0093] represents the image brightness standard value, which is the preset image brightness standard value obtained from the glaucoma differentiation detection database, and can be the average image brightness value at the image quality detection point of the preset glaucoma differentiation detection data from the historical database;

[0094] A preset image compression ratio influencing weight factor obtained from a glaucoma differentiation detection database;

[0095] is a preset run length influencing factor obtained from the glaucoma differentiation detection database, indicating the proportion of the run length in influencing the image quality evaluation index of the glaucoma differentiation detection data;

[0096] A weight factor affecting the preset image structure similarity coefficient obtained from a glaucoma differentiation detection database;

[0097] is a preset image noise impact factor obtained from the glaucoma differentiation detection database, indicating the proportion of image noise in the image quality evaluation index affecting the glaucoma differentiation detection data;

[0098] is a preset image artifact intensity influencing factor obtained from the glaucoma differentiation detection database, indicating the proportion of image artifact intensity in the image quality evaluation index affecting the glaucoma differentiation detection data.

[0099] The preset image compression ratio influence weight factor, the preset run length influence factor and the preset image structure similarity coefficient influence weight factor respectively represent the numerical value of the influence of the image compression ratio, run length and image structure similarity coefficient on the image quality assessment index of glaucoma differentiation detection data, and represent the proportion of the image compression ratio, run length and image structure similarity coefficient in the influence on the image quality assessment index of glaucoma differentiation detection data.

[0100] The preset image compression ratio influencing weight factor, the preset run length influencing factor and the preset image structure similarity coefficient influencing weight factor are obtained through mapping relationships. For example, a mapping set of image compression ratio, run length and image structure similarity coefficient and their corresponding weights is established respectively through the relationship between the image compression ratio, run length and image structure similarity coefficient in historical data and the image resolution. The corresponding preset image compression ratio influencing weight factor, preset run length influencing weight factor and preset image structure similarity coefficient influencing weight factor in the mapping set are obtained by inputting the real-time image compression ratio, run length and image structure similarity coefficient.

[0101] The preset image noise impact factor and the preset image artifact intensity impact factor respectively represent the numerical values ​​of the impact of image noise and image artifact intensity on the image quality evaluation index of glaucoma differentiation detection data, and represent the proportion of image noise and image artifact intensity in the impact on the image quality evaluation index of glaucoma differentiation detection data.

[0102] The preset image noise influence factor and the preset image artifact intensity influence factor are obtained through mapping relationships. For example, a mapping set of image noise and image artifact intensity and their corresponding weights is established respectively through the relationship between the image noise and image artifact intensity in historical data and the filter cutoff frequency. The corresponding preset image noise influence factor and preset image artifact intensity influence factor in the mapping set are obtained by inputting real-time image noise and image artifact intensity.

[0103] As the image compression ratio increases, the image quality decreases because more image information is lost. The longer the run length, the more consecutive pixel values ​​are the same, which leads to an increase in the image compression ratio. The increase in the image compression ratio leads to an increase in image noise because additional errors are introduced in the image compression process of the glaucoma differentiation detection data. The increase in the image compression ratio leads to an increase in the intensity of image artifacts because additional artifacts may be introduced in the image compression process of the glaucoma differentiation detection data. The greater the deviation of the image contrast from the standard value of the image contrast, the more obvious the image noise will be. The greater the image noise, the greater the image brightness will affect the visibility of image artifacts because image artifacts will become more obvious or more hidden as the image brightness changes. As the image brightness increases, the image artifact intensity decreases; as the image brightness decreases, the image artifact intensity increases.

[0104] There is a positive correlation between the image compression ratio and the absolute value of the image compression ratio difference and the image quality evaluation index of the glaucoma differentiation detection data. The larger the absolute value of the image compression ratio and the image compression ratio difference, the greater the compression degree of the glaucoma differentiation detection data image, which will lose some image information, thereby affecting the image quality of the glaucoma differentiation detection data, and the larger the image quality evaluation index of the glaucoma differentiation detection data; there is a positive correlation between the absolute value of the run length and the run length standard value difference and the glaucoma differentiation detection data image quality evaluation index. The larger the absolute value of the run length and the run length standard value difference, the lower the compression efficiency of the glaucoma differentiation detection data image, and the larger the image quality evaluation index of the glaucoma differentiation detection data; there is a positive correlation between the image contrast and the square of the image contrast standard value difference and the glaucoma differentiation detection data image quality evaluation index. The larger the image contrast and the square of the image contrast standard value difference, the easier it is to expose the glaucoma differentiation detection data image, affecting the image quality of the glaucoma differentiation detection data, and glaucoma The larger the image quality assessment index of the differentiation detection data is; there is a positive correlation between the square of the difference between image brightness and the standard value of image brightness and the image quality assessment index of glaucoma differentiation detection data. The larger the square of the difference between image brightness and the standard value of image brightness, the easier it is to expose the glaucoma differentiation detection data image, affecting the image quality of the glaucoma differentiation detection data, and the larger the image quality assessment index of the glaucoma differentiation detection data is; there is a positive correlation between the square of the difference between image noise and the standard value of image noise and the image quality assessment index of glaucoma differentiation detection data. The larger the square of the difference between image noise and the standard value of image noise, the greater the decline in the image quality of the glaucoma differentiation detection data, and the larger the image quality assessment index of the glaucoma differentiation detection data is; there is a positive correlation between the square of the difference between image artifact intensity and the standard value of image artifact intensity and the image quality assessment index of glaucoma differentiation detection data. The larger the square of the difference between image artifact intensity and the standard value of image artifact intensity, the greater the decline in image quality, and the larger the image quality assessment index of glaucoma differentiation detection data is.

[0105] Furthermore, the specific steps of comprehensively analyzing and obtaining the glaucoma differentiation detection data quality accuracy evaluation index are as follows: obtaining the image resolution of the preset glaucoma differentiation detection data quality accuracy detection point through image processing software; and obtaining the glaucoma differentiation detection data quality accuracy evaluation index through comprehensive analysis of the glaucoma differentiation detection data integrity evaluation index, the glaucoma differentiation detection data image quality evaluation index and the image resolution.

[0106] In this embodiment, the specific method for analyzing and obtaining the glaucoma differentiation detection data quality accuracy evaluation index is as follows:

[0107] ;

[0108] ;

[0109] The preset glaucoma differentiation test data quality accuracy test points are numbered in sequence. Indicates the number of the test point for the quality accuracy of the glaucoma differentiation test data. , The total number of test points indicating the quality accuracy of the glaucoma differentiation test data;

[0110] represents the accuracy assessment index of glaucoma differentiation test data quality;

[0111] Indicates The glaucoma differentiation test data integrity assessment index for each glaucoma differentiation test data integrity test point;

[0112] Indicates Image resolution at each glaucoma differentiation test data quality accuracy test point;

[0113] represents an image resolution threshold, which is a preset image resolution threshold obtained from a glaucoma differentiation detection database, and may be an average image resolution value at a preset glaucoma differentiation detection data quality accuracy detection point from a historical database;

[0114] Indicates The glaucoma differentiation detection data image quality assessment index of each glaucoma differentiation detection data image quality detection point;

[0115] is the influencing weight factor of the preset glaucoma differentiation detection data integrity assessment index obtained from the glaucoma differentiation detection database;

[0116] A preset image resolution impact weight factor obtained from a glaucoma differentiation detection database;

[0117] It is the influencing weight factor of the preset glaucoma differentiation detection data image quality assessment index obtained from the glaucoma differentiation detection database.

[0118] The preset glaucoma differentiation detection data integrity assessment index influencing weight factor, the preset image resolution influencing weight factor and the preset glaucoma differentiation detection data image quality assessment index influencing weight factor respectively represent the numerical values ​​of the influence of the glaucoma differentiation detection data integrity assessment index, image resolution and glaucoma differentiation detection data image quality assessment index on the glaucoma differentiation detection data quality accuracy assessment index, and represent the proportion of the glaucoma differentiation detection data integrity assessment index, image resolution and glaucoma differentiation detection data image quality assessment index in the influence on the glaucoma differentiation detection data quality accuracy assessment index.

[0119] The preset glaucoma differentiation detection data integrity assessment index influencing weight factors, the preset image resolution influencing weight factors and the preset glaucoma differentiation detection data image quality assessment index influencing weight factors are obtained through mapping relationships. For example, a mapping set of glaucoma differentiation detection data integrity assessment index, image resolution and glaucoma differentiation detection data image quality assessment index and their corresponding weights is established respectively through the relationship between the glaucoma differentiation detection data integrity assessment index, image resolution and glaucoma differentiation detection data image quality assessment index and the signal-to-noise ratio in the historical data. The preset glaucoma differentiation detection data integrity assessment index influencing weight factors, preset image resolution influencing weight factors and preset glaucoma differentiation detection data image quality assessment index influencing weight factors corresponding to the mapping set are obtained by inputting the real-time glaucoma differentiation detection data integrity assessment index, image resolution and glaucoma differentiation detection data image quality assessment index.

[0120] Table 1 is an example table of the quality accuracy evaluation index of glaucoma differentiation detection data. The example parameters in Table 1 only take the parameters under one glaucoma differentiation detection data quality accuracy detection point for example. The weight factor Set to 0.4, weight factor Set to 0.3, weight factor Set to 0.3, The image resolution threshold is 2, as shown in Table 1.

[0121] Table 1 Example table of glaucoma differentiation detection data quality accuracy evaluation index:

[0122] ;

[0123] Image resolution refers to the number of pixels in an image, which affects the clarity and details of the image. The higher the image resolution, the more details the image contains, providing more accurate and detailed image information, and the larger the glaucoma differentiation detection data integrity assessment index; the higher the image resolution, the higher the image structure similarity and peak signal-to-noise ratio, and the smaller the glaucoma differentiation detection data image quality assessment index; the glaucoma differentiation detection data integrity assessment index reflects the integrity and reliability of the glaucoma detection data. The larger the glaucoma differentiation detection data integrity assessment index, the smaller the glaucoma differentiation detection data image quality assessment index.

[0124] It can be seen from Table 1 that the larger the glaucoma differentiation detection data integrity assessment index is, the more complete and comprehensive the information provided by the glaucoma differentiation detection data is, and there is a positive correlation between the glaucoma differentiation detection data integrity assessment index and the glaucoma differentiation detection data quality accuracy assessment index, and the larger the glaucoma differentiation detection data quality accuracy assessment index is; the higher the image resolution is, the more detailed optic nerve and retinal structure information can be provided, which may improve the accuracy of glaucoma detection data, and there is a positive correlation between the image resolution and the glaucoma differentiation detection data quality accuracy assessment index, and the larger the glaucoma differentiation detection data quality accuracy assessment index is; the larger the glaucoma differentiation detection data image quality assessment index is, the greater the impact of image quality on the accuracy of glaucoma diagnosis, and there is a negative correlation between the glaucoma differentiation detection data image quality assessment index and the glaucoma differentiation detection data quality accuracy assessment index, and the smaller the glaucoma differentiation detection data quality accuracy assessment index is.

[0125] like Figure 2 As shown, it is a schematic diagram of the index function for evaluating the quality accuracy of glaucoma differentiation detection data provided in the embodiment of the present application; x is the positive semi-axis of the horizontal axis, y is the positive semi-axis of the vertical axis, and the weight factor Set to 0.4, weight factor Set to 0.3, weight factor Set to 0.3, The image resolution threshold is 2.

[0126] Curve a indicates that if the image resolution is set to a fixed value of 1, the image quality evaluation index of the glaucoma differentiation detection data is set to a fixed value of 1, the integrity evaluation index of the glaucoma differentiation detection data is x, and the quality accuracy evaluation index of the glaucoma differentiation detection data is y, the quality accuracy evaluation index of the glaucoma differentiation detection data increases with the increase of the integrity evaluation index of the glaucoma differentiation detection data.

[0127] Furthermore, the specific steps for optimizing and adjusting the method for integrity of glaucoma differentiation detection data are as follows: if the glaucoma differentiation detection data integrity assessment index is greater than or equal to the first threshold value of the glaucoma differentiation detection data integrity assessment index, there is no need to optimize and adjust the method for integrity of glaucoma differentiation detection data; if the glaucoma differentiation detection data integrity assessment index is lower than the first threshold value of the glaucoma differentiation detection data integrity assessment index, then the difference between the glaucoma differentiation detection data integrity assessment index and the first threshold value of the glaucoma differentiation detection data integrity assessment index is used to match the corresponding adjustment plan in the glaucoma differentiation detection database.

[0128] In this embodiment, assuming that the glaucoma differentiation detection data integrity assessment index is 1, the first threshold of the glaucoma differentiation detection data integrity assessment index obtained from the glaucoma differentiation detection database is 2, and the corresponding difference is -1, then the adjustment scheme corresponding to the difference between the glaucoma differentiation detection data integrity assessment index and the first threshold of the glaucoma differentiation detection data integrity assessment index obtained from the glaucoma differentiation detection database is -1 is as follows: the high-pass filter can be used to enhance the high-frequency information in the image, such as edges and details, which can be achieved by applying a high-pass filter operator, such as using the Sobel operator; the edge features in the image can be enhanced by an edge-preserving filter, while minimizing interference with other parts of the image, which can be achieved by applying an edge-preserving filter operator, such as using the Canny edge detection, and the key features in the image include edges, corners, texture analysis, color, etc., so as to improve the glaucoma differentiation detection data integrity assessment index.

[0129] Furthermore, the specific steps of optimizing and adjusting the method for image quality of glaucoma differentiation detection data are as follows: if the image quality assessment index of glaucoma differentiation detection data is lower than or equal to the second threshold value of the image quality assessment index of glaucoma differentiation detection data, there is no need to optimize and adjust the method for image quality of glaucoma differentiation detection data; if the image quality assessment index of glaucoma differentiation detection data is greater than the second threshold value of the image quality assessment index of glaucoma differentiation detection data, then the difference between the image quality assessment index of glaucoma differentiation detection data and the second threshold value of the image quality assessment index of glaucoma differentiation detection data is used to match the corresponding adjustment plan in the glaucoma differentiation detection database.

[0130] In this embodiment, assuming that the image quality assessment index of glaucoma differentiation detection data is 3, the second threshold of the image quality assessment index of glaucoma differentiation detection data obtained from the glaucoma differentiation detection database is 1, and the corresponding difference is 2, then the adjustment scheme corresponding to the difference between the image quality assessment index of glaucoma differentiation detection data and the second threshold of the image quality assessment index of glaucoma differentiation detection data obtained from the glaucoma differentiation detection database is 2 is used. The adjustment scheme is: perform wavelet transform on the image. Wavelet transform is a method of decomposing an image into components of different scales and positions. In this example, db1 wavelet (Daubechie s wavelet) is used for transformation. Wavelet transform decomposes the image into approximate coefficients and detail coefficients. The approximate coefficients represent the low-frequency components of the image, while the detail coefficients represent the high-frequency components of the image. In this example, the decomposition to the second level is chosen. In order to remove noise, a threshold can be applied to the detail coefficient. In this example, the absolute values ​​of all detail coefficients are limited to within 0.2, which means that if the absolute value of the detail coefficient is greater than 0.2, it will be clipped to 0.2. The inverse wavelet transform is used to reconstruct the image. The reconstructed image will contain the low-frequency and high-frequency components after noise is removed, so as to reduce the image quality evaluation index of glaucoma differentiation detection data.

[0131] Furthermore, the specific steps of optimizing and adjusting the method for glaucoma differentiation detection data quality accuracy are as follows: extracting the comprehensive threshold of the glaucoma differentiation detection data quality accuracy assessment index in the glaucoma differentiation detection database, and comparing the glaucoma differentiation detection data quality accuracy assessment index with the comprehensive threshold of the glaucoma differentiation detection data quality accuracy assessment index; if the glaucoma differentiation detection data quality accuracy assessment index is greater than or equal to the comprehensive threshold of the glaucoma differentiation detection data quality accuracy assessment index, there is no need to optimize and adjust the glaucoma differentiation detection data quality accuracy.

[0132] Furthermore, the method for optimizing and adjusting the quality accuracy of glaucoma differentiation detection data also includes: if the glaucoma differentiation detection data quality accuracy assessment index is lower than the comprehensive threshold of the glaucoma differentiation detection data quality accuracy assessment index, then matching the corresponding adjustment plan in the glaucoma differentiation detection database through the difference between the glaucoma differentiation detection data quality accuracy assessment index and the comprehensive threshold of the glaucoma differentiation detection data quality accuracy assessment index.

[0133] In this embodiment, assuming that the glaucoma differentiation detection data quality accuracy evaluation index is 2.5, the comprehensive threshold of the glaucoma differentiation detection data quality accuracy evaluation index obtained from the glaucoma differentiation detection database is 3, and the corresponding difference is -0.5, then the adjustment scheme corresponding to the difference between the glaucoma differentiation detection data quality accuracy evaluation index and the comprehensive threshold of the glaucoma differentiation detection data quality accuracy evaluation index is -0.5 obtained from the glaucoma differentiation detection database. The adjustment scheme is: select an initial filter parameter, for example, the standard deviation of the Gaussian filter is 1.0, apply the Gaussian filter to the image, remove noise, evaluate the filtering effect by visual inspection or quantitative indicators such as signal-to-noise ratio, structural similarity index, etc., adjust the filter parameters according to the filtering effect, for example, if the image is still noisy, increase the standard deviation of the Gaussian filter, and repeat the above steps until the glaucoma differentiation detection data The quality accuracy evaluation index is greater than or equal to the comprehensive threshold of the quality accuracy evaluation index of the glaucoma differentiation detection data, and a preset number of iterations is reached. The above process is: a glaucoma detection image containing noise, the first iteration: apply a Gaussian filter with a standard deviation of 1.0, evaluate the filtering effect, and find that the image still has noise, but the edges and details are blurred; the second iteration: based on the results of the first iteration, increase the standard deviation of the Gaussian filter to 1.5, apply the filter again, and find that the noise has been reduced, but the edges and details are still not clear enough; the third iteration: based on the results of the second iteration, decide to try to use a bilateral filter to better preserve the edges and details, apply a bilateral filter, and adjust the parameters to balance noise removal and edge retention; continue to iterate: adjust the filter parameters each time according to the previous results until the evaluation index reaches or exceeds the comprehensive threshold, or reaches a preset number of iterations.

[0134] like Figure 3As shown, it is a structural schematic diagram of a glaucoma differentiation detection system based on multimodal fusion provided in an embodiment of the present application. The glaucoma differentiation detection system based on multimodal fusion provided in an embodiment of the present application includes: a module for collecting glaucoma differentiation detection data, a module for analyzing glaucoma differentiation detection data, a comprehensive analysis module and a module for optimizing and adjusting the quality accuracy of glaucoma differentiation detection data: a module for collecting glaucoma differentiation detection data: used for collecting and processing glaucoma differentiation detection data; a module for analyzing glaucoma differentiation detection data: used for analyzing glaucoma differentiation detection data to obtain a glaucoma differentiation detection data integrity evaluation index and a glaucoma differentiation detection data image quality evaluation index; a comprehensive analysis module: used for comprehensively analyzing the obtained To the glaucoma differentiation detection data quality accuracy evaluation index; optimize and adjust the glaucoma differentiation detection data quality accuracy method module: used to compare and analyze the glaucoma differentiation detection data integrity evaluation index with the first threshold of the glaucoma differentiation detection data integrity evaluation index, and optimize and adjust the glaucoma differentiation detection data integrity method; compare and analyze the glaucoma differentiation detection data image quality evaluation index with the second threshold of the glaucoma differentiation detection data image quality evaluation index, and optimize and adjust the glaucoma differentiation detection data image quality method; compare and analyze the glaucoma differentiation detection data quality accuracy evaluation index with the comprehensive threshold of the glaucoma differentiation detection data quality accuracy evaluation index, and optimize and adjust the glaucoma differentiation detection data quality accuracy method.

[0135] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0136] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0137] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0138] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0139] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0140] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A glaucoma differentiation detection method based on multimodal fusion, characterized in that: The following steps are involved: Collect and process glaucoma differentiation test data; Analyzing the glaucoma differentiation detection data to obtain a glaucoma differentiation detection data integrity evaluation index and a glaucoma differentiation detection data image quality evaluation index; Comprehensive analysis resulted in the quality accuracy assessment index of glaucoma differentiation detection data; Compare and analyze the glaucoma differentiation detection data integrity assessment index and the first threshold of the glaucoma differentiation detection data integrity assessment index, and optimize and adjust the glaucoma differentiation detection data integrity method; Compare and analyze the glaucoma differentiation detection data image quality assessment index and the second threshold of the glaucoma differentiation detection data image quality assessment index, and optimize and adjust the glaucoma differentiation detection data image quality method; Compare and analyze the glaucoma differentiation test data quality accuracy evaluation index and the glaucoma differentiation test data quality accuracy evaluation index comprehensive threshold, and optimize and adjust the glaucoma differentiation test data quality accuracy method; The specific steps of the method for optimizing and adjusting the integrity of glaucoma differentiation detection data are as follows: If the glaucoma differentiation detection data integrity assessment index is greater than or equal to the first threshold value of the glaucoma differentiation detection data integrity assessment index, there is no need to optimize and adjust the glaucoma differentiation detection data integrity method; If the glaucoma differentiation detection data integrity assessment index is lower than the first threshold value of the glaucoma differentiation detection data integrity assessment index, then the adjustment scheme corresponding to it in the glaucoma differentiation detection database is matched by the difference between the glaucoma differentiation detection data integrity assessment index and the first threshold value of the glaucoma differentiation detection data integrity assessment index; The specific adjustment scheme is to enhance the high-frequency information in the image of glaucoma differentiation detection data through a high-pass filter; enhance the edge features in the image through an edge-preserving filter; enhance the key features in the glaucoma differentiation detection image through an edge-preserving filter; The specific steps of the method for optimizing and adjusting the image quality of glaucoma differentiation detection data are as follows: If the glaucoma differentiation detection data image quality assessment index is lower than or equal to the glaucoma differentiation detection data image quality assessment index second threshold value, then there is no need to optimize and adjust the glaucoma differentiation detection data image quality method; If the glaucoma differentiation detection data image quality assessment index is greater than the glaucoma differentiation detection data image quality assessment index second threshold value, then the adjustment scheme corresponding to it in the glaucoma differentiation detection database is matched by the difference between the glaucoma differentiation detection data image quality assessment index and the glaucoma differentiation detection data image quality assessment index second threshold value; The specific adjustment scheme is to perform wavelet transform on the image of glaucoma differentiation detection data to remove noise and retain image features; The method for optimizing and adjusting the quality accuracy of glaucoma differentiation detection data also includes: If the glaucoma differentiation detection data quality accuracy assessment index is lower than the glaucoma differentiation detection data quality accuracy assessment index comprehensive threshold, then the corresponding adjustment scheme in the glaucoma differentiation detection database is matched by the difference between the glaucoma differentiation detection data quality accuracy assessment index and the glaucoma differentiation detection data quality accuracy assessment index comprehensive threshold; The specific adjustment scheme is to apply an iterative filtering operation to the image of the glaucoma differentiation detection data through an iterative filter, select initial filter parameters, evaluate the filtering effect through visual inspection or quantitative indicators, adjust the filter parameters according to the filtering effect, and repeat the iterative filtering operation steps until the glaucoma differentiation detection data quality accuracy assessment index is greater than or equal to the glaucoma differentiation detection data quality accuracy assessment index comprehensive threshold.

2. The glaucoma differentiation detection method based on multimodal fusion as claimed in claim 1, characterized in that: The specific steps of collecting and processing glaucoma differentiation detection data are: Collecting raw data of glaucoma differentiation detection through sensor equipment; The raw data of glaucoma differentiation detection are cleaned and denoised to obtain glaucoma differentiation detection data, wherein the glaucoma differentiation detection data includes glaucoma differentiation detection data integrity data and glaucoma differentiation detection data image quality data.

3. The glaucoma differentiation detection method based on multimodal fusion as claimed in claim 1, characterized in that: The specific steps of obtaining the glaucoma differentiation detection data integrity evaluation index are: The entropy value of the glaucoma differentiation detection data of the preset glaucoma differentiation detection data integrity detection point is obtained by image processing software; The run length of the preset glaucoma differentiation detection data integrity detection point is obtained by image compression software; The maximum value of the peak signal intensity of the preset glaucoma differentiation detection data integrity detection point is obtained by image processing software; The minimum value of the peak signal intensity of the preset glaucoma differentiation detection data integrity detection point is obtained by image processing software; The image noise of the preset glaucoma differentiation detection data integrity detection point is obtained by using a noise analysis tool; The image artifact intensity of the preset glaucoma differentiation detection data integrity detection point is obtained by image processing software; Obtaining the ambient light angle of the preset glaucoma differentiation detection data integrity detection point through the light measurement device; The glaucoma differentiation detection data integrity data includes glaucoma differentiation detection data entropy value, run length, peak signal intensity maximum value, peak signal intensity minimum value, image noise, image artifact intensity and ambient light angle; The glaucoma differentiation detection data integrity assessment index was obtained based on the glaucoma differentiation detection data integrity data analysis.

4. The glaucoma differentiation detection method based on multimodal fusion as claimed in claim 1, characterized in that: The specific steps of obtaining the image quality evaluation index of glaucoma differentiation detection data are as follows: The image compression ratio of the preset glaucoma differentiation detection data image quality detection point is obtained by image compression software; The run length of the preset glaucoma differentiation detection data image quality detection point is obtained by image compression software; The image contrast of the preset glaucoma differentiation detection data image quality detection point is obtained by image processing software; The image brightness of the preset glaucoma differentiation detection data image quality detection point is obtained by image processing software; The image noise of the preset glaucoma differentiation detection data image quality detection point is obtained by image processing software; The image artifact intensity of the preset glaucoma differentiation detection data image quality detection point is obtained by image processing software; The image quality data of the glaucoma differentiation detection data includes image compression ratio, run length, image contrast, image brightness, image noise, and image artifact intensity; The image quality evaluation index of glaucoma differentiation detection data is obtained based on the image quality data analysis of glaucoma differentiation detection data.

5. The glaucoma differentiation and detection method based on multimodal fusion as claimed in claim 1, characterized in that: The specific steps of comprehensively analyzing and obtaining the quality accuracy evaluation index of glaucoma differentiation detection data are as follows: The image resolution of the preset glaucoma differentiation detection data quality accuracy detection point is obtained through image processing software; The glaucoma differentiation detection data quality accuracy assessment index was obtained through comprehensive analysis of the glaucoma differentiation detection data integrity assessment index, the glaucoma differentiation detection data image quality assessment index and the image resolution.

6. The glaucoma differentiation and detection method based on multimodal fusion as claimed in claim 1, characterized in that: The specific steps of the method for optimizing and adjusting the quality accuracy of glaucoma differentiation detection data are as follows: Extracting the comprehensive threshold of the glaucoma differentiation detection data quality accuracy assessment index from the glaucoma differentiation detection database, and comparing the glaucoma differentiation detection data quality accuracy assessment index with the comprehensive threshold of the glaucoma differentiation detection data quality accuracy assessment index; If the glaucoma differentiation detection data quality accuracy assessment index is greater than or equal to the glaucoma differentiation detection data quality accuracy assessment index comprehensive threshold, there is no need to optimize and adjust the glaucoma differentiation detection data quality accuracy.

7. A system using the glaucoma differentiation and detection method based on multimodal fusion as described in any one of claims 1 to 6, characterized in that: It includes a module for collecting glaucoma differentiation detection data, a module for analyzing glaucoma differentiation detection data, a comprehensive analysis module, and a module for optimizing and adjusting the quality and accuracy of glaucoma differentiation detection data: Glaucoma differentiation detection data collection module: used to collect and process glaucoma differentiation detection data; A module for analyzing glaucoma differentiation detection data: used to analyze glaucoma differentiation detection data, obtain a glaucoma differentiation detection data integrity evaluation index and obtain a glaucoma differentiation detection data image quality evaluation index; Comprehensive analysis module: used for comprehensive analysis to obtain the quality accuracy evaluation index of glaucoma differentiation detection data; Module for optimizing and adjusting the quality accuracy of glaucoma differentiation detection data: used to compare and analyze the glaucoma differentiation detection data integrity assessment index with the first threshold of the glaucoma differentiation detection data integrity assessment index, and optimize and adjust the glaucoma differentiation detection data integrity method; Compare and analyze the glaucoma differentiation detection data image quality assessment index and the second threshold of the glaucoma differentiation detection data image quality assessment index, and optimize and adjust the glaucoma differentiation detection data image quality method; compare and analyze the glaucoma differentiation detection data quality accuracy assessment index and the comprehensive threshold of the glaucoma differentiation detection data quality accuracy assessment index, and optimize and adjust the glaucoma differentiation detection data quality accuracy method.

Images