AI-based CBA fluorescence immunoassay judgment method and system and storage medium

Abstract

The invention relates to the technical field of medical diagnosis, in particular to an AI-based CBA fluorescence immunoassay judgment method and system and a storage medium, and the AI-based CBA fluorescence immunoassay judgment method comprises the following steps: based on fluorescence immunoassay image data, analyzing and identifying the illumination intensity and vibration error of an image, adjusting exposure parameters and a stabilization algorithm to eliminate errors and noisy points, and obtaining a fluorescence immunoassay result; and generating an image correction processing record. The method has the beneficial effects that the illumination intensity and the vibration error of the fluorescence immune image data are analyzed and adjusted, the noisy points and the error of the image are eliminated, and the definition of a plurality of key biomarkers in the image is optimized by combining image edge information enhancement, contrast adjustment and sharpening processing; the method improves the accuracy and efficiency of visual recognition of the antibody, improves the judgment capability of the type and characteristics of the antibody by calculating the spatial distribution and relative brightness of the antibody in the image, rapidly diagnoses and classifies diseases, and improves the efficiency and speed of processing batch samples.

Description

Technical Field

[0001] The present invention relates to the field of medical diagnosis technology, and in particular to an AI-based CBA fluorescence immunoassay method, system and storage medium. Background Art

[0002] The field of medical diagnostic technology aims to use a variety of tools and methods to detect and evaluate health conditions, assist in the diagnosis and treatment of diseases, and process batch medical data through biomarker identification, molecular imaging, and biochemical analysis, combined with artificial intelligence and machine learning, to improve the accuracy and efficiency of diagnosis, realize automated image analysis and interpretation, improve processing efficiency and accuracy, reduce the burden on medical professionals, help doctors predict disease progression, optimize treatment plans, and provide patients with personalized medical services.

[0003] Among them, the AI-based CBA fluorescent immunoassay judgment method, by combining artificial intelligence technology, aims to improve the accuracy and speed of fluorescent immunoassay in a computer-assisted manner, use AI algorithms to identify and analyze fluorescently labeled images, accurately quantify the expression of biomarkers, and provide fast and accurate pathological analysis to help doctors make more accurate diagnostic decisions. Combined with automated processing and analysis of batch sample data, it provides key diagnostic information and provides a data basis for disease management and the formulation of treatment plans.

[0004] Traditional fluorescent immunoassay methods are limited in terms of antibody recognition due to the limitations of image quality and analysis methods. As a result, uneven lighting and equipment vibration affect the image quality, which reduces the accuracy of antibody recognition. It lacks sufficient precision when processing complex fluorescence patterns and has difficulty distinguishing similar antibody types. When antibodies are weakly expressed and unevenly distributed, it leads to increased errors in antibody analysis, misdiagnosis and missed diagnosis, affecting the early diagnosis and treatment of the disease. When analyzing batch sample data, the processing efficiency is poor, limiting the timeliness and efficiency of the diagnostic process. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose an AI-based CBA fluorescent immunoassay method, system and storage medium.

[0006] In order to achieve the above object, the present invention adopts the following technical scheme, the CBA fluorescence immunoassay method based on AI comprises the following steps: S1: Based on the fluorescence immunoassay data, the image light intensity and vibration errors are analyzed and identified, the exposure parameters and stabilization algorithm are adjusted to eliminate errors and noise, and the image correction processing record is generated; S2: Based on the image correction processing record, identifying and enhancing edge information in the image, combining contrast adjustment and sharpening processing, optimizing the clarity of multiple key features in the image, and generating an image feature enhancement result; S3: classifying the cell and tissue features in the image according to the image feature enhancement result, identifying multiple pathological conditions by analyzing the matching degree between the image features and known pathological characteristics, and generating a pathological condition classification result; S4: According to the pathological state classification result, by analyzing the data features in the pathological image, matching pathological feature information for multiple pathological results, and generating a diagnostic information matching result; S5: Based on the diagnostic information matching result, analyzing the fluorescence patterns of multiple antibodies in the feature-enhanced image, identifying the intensity, distribution and dynamic changes of the fluorescence of multiple antibodies, and obtaining the fluorescence pattern analysis result; S6: Based on the fluorescence pattern analysis results, the antibody types present in the image are analyzed through image recognition and biomarker database comparison, and the detection time and disease types of multiple antibodies are recorded, and the antibody type analysis record is output.

[0007] As a further scheme of the present invention, the image correction processing record includes adjusted exposure parameters, eliminated vibration errors, and reduced number of noise points; the image feature enhancement results include enhanced edge information, adjusted contrast parameters, and sharpened key features; the pathological state classification results include classified cell types, identified tissue features, and matched pathological states; the diagnostic information matching results include corresponding health risk levels, etiology-related information, and reference pathological data; the fluorescence pattern analysis results include identified antibody fluorescence intensity, calculated spatial distribution, and measured relative brightness; the antibody type analysis record includes recorded antibody type, detection time, and associated pathological type.

[0008] As a further solution of the present invention, based on the fluorescence immunoassay image data, the illumination intensity and vibration errors of the image are analyzed and identified, the exposure parameters and the stabilization algorithm are adjusted to eliminate the errors and noise, and the steps of generating the image correction processing record are specifically as follows: S101: Analyze the light intensity of the image according to the fluorescence immunoassay image data, record the brightness values ​​of multiple areas, adjust the exposure parameters to optimize the lighting effect, and generate a brightness adjustment record; S102: Analyzing image vibration errors based on the brightness adjustment record, adjusting image stabilization parameters to eliminate vibration caused by the device and optimize image stability, and generating a stability optimization record; S103: Based on the stability optimization record, adjust the filtering strength and parameters, optimize the random noise and color distortion in the image, and generate an image correction processing record.

[0009] As a further solution of the present invention, based on the image correction processing record, edge information in the image is identified and enhanced, contrast adjustment and sharpening are combined to optimize the clarity of multiple key features in the image, and the steps of generating image feature enhancement results are specifically as follows: S201: Based on the image correction processing record, edge information in the image is identified, including analyzing edge clarity and continuity to obtain edge identification results; S202: Based on the edge recognition result, adjusting the contrast of multiple regions in the image to optimize the distinguishability of key features, and generating a contrast processing result; S203: Based on the contrast processing result, adjust the image sharpening parameters to optimize the clarity of multiple key features in the image to obtain an image feature enhancement result.

[0010] As a further solution of the present invention, according to the image feature enhancement result, the cell and tissue features in the image are classified, and a plurality of pathological conditions are identified by analyzing the matching degree between the image features and the known pathological features. The specific steps of generating the pathological condition classification result are as follows: S301: Analyze the image based on the image feature enhancement result, identify the structural features of cells and tissues in the image, including cell shape, size and color, and generate a cell structure analysis result; S302: Based on the cell structure analysis result, compare the cell and tissue features in the image with the pathological feature database and calculate the matching degree to obtain the pathological feature matching result; S303: Based on the pathological feature matching result, classify the cell and tissue features in the image, and identify multiple pathological states according to the matching degree to obtain a pathological state classification result.

[0011] As a further solution of the present invention, according to the pathological state classification result, by analyzing the data features in the pathological image, matching pathological feature information for multiple pathological results, and generating the diagnostic information matching result are specifically as follows: S401: using the pathological state classification results and comparing with pathological data, performing data feature analysis on multiple pathological states, recording key data features, and generating pathological data analysis information; S402: Based on the pathological data analysis information and in combination with the pathological database, health risk assessment and etiology analysis are performed on various pathological conditions to obtain health risk analysis data; S403: According to the health risk analysis data, diagnostic information is matched for the target image, including health risk assessment results of various pathological conditions and matching degrees of various causes of disease, to obtain a diagnostic information matching result.

[0012] As a further solution of the present invention, based on the diagnostic information matching results, the steps of analyzing multiple antibody fluorescence patterns in the feature-enhanced image, identifying the intensity, distribution and dynamic changes of multiple antibody fluorescence, and obtaining the fluorescence pattern analysis results are specifically as follows: S501: Based on the diagnostic information matching result, perform spectral analysis on the fluorescence image, identify the fluorescence intensity of multiple antibodies, record the spectral data of the multiple antibodies, and generate fluorescence intensity data; S502: Based on the fluorescence intensity data, a kernel density estimation algorithm is used to calculate the spatial distribution of antibody fluorescence in the image, evaluate the uniformity and aggregation area of ​​fluorescence distribution, and obtain fluorescence spatial distribution information; S503: Based on the fluorescence spatial distribution information, analyze the dynamic changes of fluorescence, evaluate the fluorescence change trend in the time series, and obtain the fluorescence pattern analysis result.

[0013] As a further solution of the present invention, the kernel density estimation algorithm is according to the formula: ; Calculate the fluorescence density estimate, where For the evaluation point The estimated density value at is the total number of data points, For the evaluation point, For the The location of the data points, is the global bandwidth, To dynamically adjust parameters, is the spatial resolution factor, Adjust the parameters for the kernel shape, is the kernel function.

[0014] As a further solution of the present invention, based on the fluorescence pattern analysis results, the antibody types present in the image are analyzed and the detection time and disease types of multiple antibodies are recorded through image recognition and biomarker database comparison. The steps of outputting the antibody type analysis record are specifically as follows: S601: Analyze and identify spectral features of multiple antibody fluorescence images based on the fluorescence pattern analysis results, including intensity features and distribution patterns, and generate antibody spectral feature records; S602: Based on the antibody spectral feature records, use image recognition to analyze the spectral data of each antibody, and compare it with the antibody features in the known biomarker database to obtain an antibody matching result; S603: According to the antibody matching results, the detection time and associated disease types of multiple antibodies are recorded to generate an antibody type analysis record.

[0015] The AI-based CBA fluorescence immunoassay judgment system is used to execute the AI-based CBA fluorescence immunoassay judgment method, and the AI-based CBA fluorescence immunoassay judgment system includes: The exposure parameter adjustment module collects fluorescent immunoassay image data, analyzes the light intensity and vibration errors of multiple areas in the image, adjusts the exposure parameters and eliminates image errors and noise, and generates image correction results; The edge information optimization module receives the image correction result, identifies the edge information in the image, and optimizes the clarity of key features in the image by combining contrast adjustment and sharpening processing to generate key feature processing results; The pathology recognition and classification module classifies cell and tissue features based on the key feature processing results, identifies multiple pathological states matching pathological feature information, including health risks and causes, by comparing the matching degree between image features and known pathological features, and generates pathological information diagnosis records; The fluorescence pattern evaluation module analyzes the fluorescence patterns of multiple antibodies in the feature-enhanced image based on the pathological information diagnosis record, identifies the intensity, distribution and dynamic changes of the fluorescence, calculates the spatial distribution and relative brightness of the fluorescence, and generates antibody fluorescence distribution data; The detection information recording module receives the antibody fluorescence distribution data, analyzes and records the types, detection times and associated disease types of multiple antibodies in the image, and generates an antibody type analysis record.

[0016] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the AI-based CBA fluorescent immunoassay method as described above.

[0017] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, the illumination intensity and vibration error of fluorescent immunoassay data are analyzed and adjusted to eliminate image noise and errors, and the clarity of multiple key biomarkers in the image is optimized by combining image edge information enhancement, contrast adjustment and sharpening processing, thereby improving the accuracy and efficiency of visual recognition of antibodies. By calculating the spatial distribution and relative brightness of antibodies in the image, the ability to judge the type and characteristics of antibodies is improved, diseases are quickly diagnosed and classified, and the efficiency and speed of processing batch samples are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a schematic diagram of the workflow of the present invention; Figure 2 This is a detailed flow chart of S1 of the present invention; Figure 3 This is a detailed flow chart of S2 of the present invention; Figure 4 This is a detailed flow chart of S3 of the present invention; Figure 5 This is a detailed flow chart of S4 of the present invention; Figure 6 This is a detailed flow chart of S5 of the present invention; Figure 7 This is a detailed flow chart of S6 of the present invention; Figure 8 It is a system flow chart of the present invention. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0020] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0021] See also Figure 1 The present invention provides a technical solution, a CBA fluorescence immunoassay method based on AI, comprising the following steps: S1: Based on the fluorescence immunoassay data, the image light intensity and vibration errors are analyzed and identified, the exposure parameters and stabilization algorithm are adjusted to eliminate errors and noise, and the image correction processing record is generated; S2: Based on the image correction processing record, identify and enhance the edge information in the image, combine contrast adjustment and sharpening processing to optimize the clarity of multiple key features in the image, and generate image feature enhancement results; S3: Based on the image feature enhancement results, the cell and tissue features in the image are classified, and by analyzing the matching degree between the image features and the known pathological features, multiple pathological conditions are identified and the pathological condition classification results are generated; S4: According to the pathological state classification result, by analyzing the data features in the pathological image, matching the pathological feature information for the multiple pathological results, and generating the diagnostic information matching result; S5: Based on the diagnostic information matching results, analyze the fluorescence patterns of multiple antibodies in the feature-enhanced image, identify the intensity, distribution and dynamic changes of the fluorescence of multiple antibodies, and obtain the fluorescence pattern analysis results; S6: Based on the fluorescence pattern analysis results, through image recognition and biomarker database comparison, analyze the antibody types present in the image and record the detection time and disease types of multiple antibodies, and output the antibody type analysis record.

[0022] Image correction processing records include adjusted exposure parameters, eliminated vibration errors, and reduced number of noise points. Image feature enhancement results include enhanced edge information, adjusted contrast parameters, and sharpened key features. Pathological state classification results include classified cell types, identified tissue features, and matched pathological states. Diagnostic information matching results include corresponding health risk levels, etiology-related information, and referenced pathological data. Fluorescence pattern analysis results include identified antibody fluorescence intensity, calculated spatial distribution, and measured relative brightness. Antibody type analysis records include recorded antibody type, detection time, and associated pathology type.

[0023] See also Figure 2 ,Based on the fluorescence immunoassay data, the image illumination intensity and vibration errors are analyzed and identified, the exposure parameters and stabilization algorithm are adjusted to eliminate errors and noise, and the steps to generate image correction processing records are as follows: S101: Analyze the illumination intensity of the image according to the fluorescence immunoassay image data, record the brightness values ​​of multiple areas, adjust the exposure parameters to optimize the illumination effect, and generate the brightness adjustment record. In sub-step S101, based on the fluorescence immunoassay image data, the brightness value of each pixel in the image is obtained through image processing software, a brightness distribution model is constructed, the brightness value of each area is statistically analyzed, the average brightness and standard deviation are calculated, and the exposure parameters of the camera are adjusted according to the statistical results to make the image brightness evenly distributed, while reducing the contrast difference between highlight and dark areas. A dynamic adjustment algorithm is used to dynamically modify the exposure time and aperture size according to the real-time feedback brightness data. The formula is: ,in, is the adjusted brightness, is the original brightness, is the target brightness, It is an adjustment coefficient used to control the sensitivity of brightness adjustment, optimize the overall brightness of the image, and generate a brightness adjustment record.

[0024] S102: Based on the brightness adjustment record, analyzing the image vibration error, adjusting the image stabilization parameters to eliminate the vibration caused by the device and optimize the image stability, and generating the stability optimization record specifically includes: In sub-step S102, according to the brightness adjustment record, the vibration error of the image is analyzed, and the vibration mode caused by the device is identified by analyzing the time series of continuous multi-frame images, and the image stabilization parameters are adjusted using the target data. For each frame of the image, the displacement difference with the previous frame is calculated, and the sliding window technology is applied to average the multi-frame data to reduce the random error. The feedback control algorithm is used to dynamically adjust the anti-shake algorithm parameters according to the real-time vibration analysis data. The formula is: ,in, is the adjusted image stability parameter, is the original parameter, is the error change, To adjust speed parameters, eliminate vibrations caused by the device and optimize image stability, a stability optimization record is generated.

[0025] S103: Based on the stability optimization record, the filtering strength and parameters are adjusted to optimize the random noise and color distortion in the image, and the process of generating the image correction processing record is specifically as follows; In sub-step S103, based on the stability optimization record, the filter strength and parameters are adjusted to reduce random noise and color distortion in the image. The image is subjected to multi-level filtering using digital filtering technology. Each level of filtering is optimized based on the output of the previous level. A preliminary filtering strength is set, and the filtering parameters are gradually adjusted according to the image quality feedback. The formula is: ,in, is the new filter parameter, is the old filter parameter, To maintain the coefficients, control the basic rate of change of the filter parameters, is the adjustment coefficient, which is adjusted according to the difference between the target and current image quality. is the target image quality, Generates an image correction processing record for the current image quality, optimizing random noise and color distortion in the image.

[0026] See also Figure 3 Based on the image correction processing record, the edge information in the image is identified and enhanced, and the clarity of multiple key features in the image is optimized by combining contrast adjustment and sharpening processing. The specific steps for generating image feature enhancement results are as follows: S201: Based on the image correction processing record, edge information in the image is identified, including analyzing the clarity and continuity of the edge, and the process of obtaining the edge identification result is specifically as follows; In sub-step S201, based on the image correction processing record, the edge information in the image is identified, and the edge clarity and continuity of each area are analyzed using an edge detection algorithm, including calculating the gradient strength and direction of edge pixels. By constructing an edge strength map, the parts with clearer and more continuous edges are identified and marked, and the edge is clarified using an edge enhancement formula, which is: ,in, is the edge clarity after adjustment, is the original edge definition, is the target gradient threshold, is the current gradient value, It is an adjustment coefficient used to control the degree of edge definition enhancement, improve the clarity and continuity of the edges in the image, and generate edge recognition results.

[0027] S202: Based on the edge recognition result, the contrast of multiple regions in the image is adjusted to optimize the discrimination of key features, and the process of generating the contrast processing result is specifically as follows; In sub-step S202, the contrast in the image is adjusted according to the edge recognition result to optimize the discrimination of key features. The contrast of each region is adjusted independently, and the local contrast adjustment formula is used to enhance the visibility of key features. The formula is: ,in, is the adjusted contrast, is the original contrast, To maintain the coefficient, control the basic rate of contrast adjustment, To adjust the coefficient, adjust according to the difference between the target and current contrast. is the target contrast, For the current contrast, enhance the distinction of multiple key features in the image and generate a contrast processing result.

[0028] S203: Based on the contrast processing result, the image sharpening parameters are adjusted to optimize the clarity of multiple key features in the image, and the process of obtaining the image feature enhancement result is specifically as follows; In sub-step S203, based on the contrast processing result, the image sharpening parameters are adjusted to optimize the clarity of key features in the image. According to the output of the contrast adjustment, the area that needs additional sharpening processing is determined, and the edge strength and detail performance of the feature are adjusted using the image sharpening formula, which is: ,in, is the adjusted sharpening strength, is the original sharpening strength, is an adjustment factor used to adjust the sharpening strength according to the clarity requirements. is the target sharpening strength, The current sharpening strength is used to improve the clarity of multiple key features in the image and obtain image feature enhancement results.

[0029] See also Figure 4 , according to the image feature enhancement results, the cell and tissue features in the image are classified, and multiple pathological states are identified by analyzing the matching degree between the image features and the known pathological features. The specific steps for generating the pathological state classification results are as follows: S301: Based on the image feature enhancement result, the image is analyzed to identify the structural features of cells and tissues in the image, including cell shape, size and color, and the specific process of generating the cell structure analysis result is as follows; In sub-step S301, based on the image feature enhancement results, the structural feature analysis of cells and tissues is performed, including using image processing technology to identify and measure the shape, size and color of cells, using morphological analysis tools to perform edge detection and shape matching on cell contours, and measuring cell size. Color analysis is performed through a color segmentation algorithm to distinguish different color features of cell tissues. The formula is: ,in, is the adjusted cell structure characteristic data, is the original data obtained by measurement, is the ideal or expected data, The adjustment coefficient is used to adjust the measurement results according to the expected value. By recording the structural characteristics of cells and tissues, the cell structure analysis results are generated.

[0030] S302: Based on the cell structure analysis result, the cell and tissue features in the image are compared with the pathological feature database and the matching degree is calculated to obtain the pathological feature matching result. The specific process is as follows; In sub-step S302, based on the cell structure analysis results, the cell and tissue features in the image are compared with the pathological feature database, and the matching degree is calculated by calculating the similarity between the image features and multiple known pathological features in the database. The similarity analysis algorithm is used, and the formula is: ,in, Score the match. For the The weight of the feature, is the eigenvalue in the image, is the corresponding pathological feature value in the database, The number of features is used to analyze the matching degree between the features in the image and the known pathological conditions, and generate pathological feature matching results.

[0031] S303: Based on the pathological feature matching results, the cell and tissue features in the image are classified, and multiple pathological states are identified according to the matching degree. The specific process of obtaining the pathological state classification result is as follows; In sub-step S303, based on the pathological feature matching results, the cell and tissue features in the image are classified, and multiple pathological states are identified according to the matching degree. The matching degree score is calculated using a statistical method, and the score is calculated for each pathological state. The formula is: ,in, is the calculation score of the classification result, For the The matching degree of the features, For the The target pathological state score of each feature, The pathological states of cells and tissues are classified based on the number of features and the classification results of the pathological states are obtained.

[0032] See also Figure 5 According to the pathological state classification results, by analyzing the data features in the pathological image, the pathological feature information is matched for multiple pathological results, and the steps of generating the diagnostic information matching results are as follows: S401: Using the pathological status classification results and comparing the pathological data, performing data feature analysis on various pathological statuses, recording key data features, and generating pathological data analysis information are specifically as follows; In sub-step S401, based on the pathological state classification results and compared with the pathological data, data feature analysis is performed to identify key biomarkers and data features by analyzing the data patterns under various pathological states, and to extract and record various key data features using statistical methods. The formula is: ,in, is the average key data characteristic value obtained, For the Frequency data of pathological conditions, For the The values ​​of key biomarkers, To determine the number of features to be analyzed, key data features of multiple pathological conditions are recorded and analyzed to generate pathological data analysis information.

[0033] S402: Based on the pathological data analysis information and in combination with the pathological database, health risk assessment and etiology analysis are performed on various pathological conditions to obtain health risk analysis data. Specifically, the process is as follows; In sub-step S402, based on the pathological data analysis information and combined with the pathological database, health risk assessment and etiology analysis are performed, and the health risk level of each pathological state is calculated using model comparison and risk scoring algorithms to assess the likelihood and severity of the disease. The formula is: ,in, Score health risks, For the The health risk weight of each pathological state is is the predicted causal probability of the state, The health risk analysis data are generated by averaging the causal probabilities and providing health risk assessment for each pathological state.

[0034] S403: According to the health risk analysis data, the target image is matched with diagnostic information, including health risk assessment results of various pathological conditions and matching degrees of various causes of disease, and the specific process of obtaining the diagnostic information matching result is as follows; In sub-step S403, the health risk analysis data is used to match the target image with diagnostic information, and the matching degree of the pathological state in the image with the known cause database is evaluated and compared through pattern recognition and matching algorithms. The formula is: ,in, The highest diagnostic match is For the The matching score of each pathological state, To verify the score, characterize the reliability of the matching results, is the threshold used to determine the acceptance level of the match, analyze the health risk assessment results of various pathological states in the image and the matching degree of various causes, and generate diagnostic information matching results.

[0035] See also Figure 6 Based on the diagnostic information matching results, the steps of analyzing the fluorescence patterns of multiple antibodies in the feature-enhanced image, identifying the intensity, distribution and dynamic changes of the fluorescence of multiple antibodies, and obtaining the fluorescence pattern analysis results are as follows: S501: Based on the diagnostic information matching result, the fluorescence image is spectrally analyzed to identify the fluorescence intensity of multiple antibodies, and the spectral data of multiple antibodies are recorded to generate fluorescence intensity data. Specifically, the process is as follows; In sub-step S501, the diagnostic information matching result is used to perform spectral analysis on the fluorescence image, including identifying and recording the intensity of the fluorescence of multiple antibodies, measuring the fluorescence emitted by each antibody, and recording the fluorescence spectrum characteristics through the data captured by the integrating sphere and the spectrometer. The formula is: ,in, is the newly calculated fluorescence intensity data, For the The spectral signal intensity of the antibodies, is the light intensity in spectral measurement, To measure the number of antibody species, the spectral data of multiple antibodies were analyzed to generate fluorescence intensity data.

[0036] S502: Based on the fluorescence intensity data, a kernel density estimation algorithm is used to calculate the spatial distribution of antibody fluorescence in the image, and the uniformity and aggregation area of ​​the fluorescence distribution are evaluated to obtain the fluorescence spatial distribution information. Specifically, the process is as follows; In sub-step S502, the spatial distribution of antibody fluorescence in the image is calculated based on the fluorescence intensity data, including evaluating the uniformity and aggregation area of ​​the fluorescence distribution, and using image processing software to perform spatial analysis and quantitative evaluation to evaluate the uniformity of the fluorescence distribution and the aggregation of the target area, and generate fluorescence spatial distribution information.

[0037] Kernel density estimation algorithm, according to the formula: ; Calculate the fluorescence density estimate, where is the kernel density estimation function, which is used to represent the evaluation point The estimated density value at the position represents the antibody fluorescence density at that position in the entire image and is used to analyze and visualize the spatial distribution of antibodies. is the total number of data points, used to represent the total amount of antibody spots in the processed image. is the evaluation point, which is used to indicate the specific location in the image being evaluated. For the The location of the data point is used to represent the The specific location of the antibody is the global bandwidth, which is used to control the degree of smoothing in the kernel density estimation. To dynamically adjust parameters, according to the position The local density changes are adjusted to enhance the sensitivity to high variation areas. is the spatial resolution factor, which is used to optimize the kernel density estimation of the local area by adjusting the relationship between bandwidth and spatial resolution. Adjust the parameters of the kernel function shape to adjust the steepness of the kernel function and enhance or weaken the influence of local features. is the kernel function, which is used to calculate the density contribution between two points. The kernel function shape and parameters are given by and Adjustment.

[0038] The specific execution process of the formula is as follows: Obtain the location of each antibody through fluorescence image data According to the antibody distribution and intensity changes, the model is adjusted using local density to obtain dynamic parameters , while determining the spatial resolution factor and kernel function shape adjustment parameters , substitute the parameters into the formula and calculate each evaluation point The spatial distribution of fluorescence was used to assess the distribution and aggregation of antibodies in the image.

[0039] S503: Based on the fluorescence spatial distribution information, the dynamic change of fluorescence is analyzed, the fluorescence change trend in the time series is evaluated, and the specific process of obtaining the fluorescence pattern analysis result is as follows; In sub-step S503, based on the fluorescence spatial distribution information, the dynamic changes of fluorescence are analyzed, including evaluating the fluorescence change trend under the time series, and using time series analysis to track the change pattern of fluorescence intensity. The formula is: ,in, The results of the dynamic change trend of fluorescence are shown in Figure 2. For the Antibodies at time The rate of change within the interval, This analysis method is used to evaluate and predict the time-dependent fluorescence variation pattern for each time increment within the observation period, thereby obtaining the fluorescence pattern analysis result.

[0040] See also Figure 7 Based on the fluorescence pattern analysis results, the antibody types present in the image are analyzed through image recognition and biomarker database comparison, and the detection time and disease types of multiple antibodies are recorded. The specific steps for outputting the antibody type analysis record are as follows: S601: Based on the fluorescence pattern analysis results, the spectral characteristics of multiple antibody fluorescence images are analyzed and identified, including intensity characteristics and distribution patterns, and the process of generating antibody spectral characteristic records is specifically as follows; In sub-step S601, according to the fluorescence pattern analysis results, the spectral characteristics of multiple antibody fluorescence images are analyzed and identified, including analyzing the intensity characteristics and distribution patterns of antibody fluorescence, and using quantitative spectral analysis technology to record the spectral intensity and spatial distribution data of each antibody. The formula is: ,in, For antibody spectral feature recording, For the The spectral intensity of the antibodies, is the corresponding distribution mode index, The number of antibody types, the spectral characteristics of multiple antibodies are recorded, and the antibody spectral characteristic record is generated.

[0041] S602: Based on the antibody spectral feature record, the spectral data of each antibody is analyzed using image recognition, and compared with the antibody features in the known biomarker database to obtain the antibody matching result. The specific process is as follows; In sub-step S602, the spectral data of each antibody is analyzed by image recognition technology using the antibody spectral feature records, and the various data are compared with the antibody features in the known biomarker database to identify and verify the antibody type. The formula is: ,in, For antibody matching results, For the The correlation weights between the spectral data of the antibodies and the database features, is the measured data value, The reference data value in the database ensures the accuracy of antibody feature matching and identification, and obtains antibody matching results.

[0042] S603: According to the antibody matching results, the detection time and associated disease types of multiple antibodies are recorded, and the specific process of generating antibody type analysis records is as follows; In sub-step S603, based on the antibody matching results, the detection time of multiple antibodies and the associated disease types are recorded to understand the function of the antibodies and the correlation with disease diagnosis. The formula is: ,in, For antibody type analysis records, For the Antibody detection time, For the associated disease type indicator, the time of antibody detection and disease type information are recorded to generate an antibody type analysis record.

[0043] See also Figure 8 The AI-based CBA fluorescence immune judgment system is used to execute the AI-based CBA fluorescence immune judgment method. The AI-based CBA fluorescence immune judgment system includes: The exposure parameter adjustment module collects fluorescent immunoassay image data, analyzes the light intensity and vibration errors of multiple areas in the image, adjusts the exposure parameters and eliminates image errors and noise, and generates image correction results; The edge information optimization module receives the image correction result, identifies the edge information in the image, and optimizes the clarity of key features in the image by combining contrast adjustment and sharpening processing to generate key feature processing results; The pathology recognition and classification module classifies cell and tissue features based on the key feature processing results, identifies multiple pathological states and matches pathological feature information, including health risks and causes, by comparing the matching degree between image features and known pathological features, and generates pathological information diagnosis records; The fluorescence pattern evaluation module analyzes the fluorescence patterns of multiple antibodies in the feature-enhanced image based on the pathological information diagnosis record, identifies the intensity, distribution and dynamic changes of fluorescence, calculates the spatial distribution and relative brightness of fluorescence, and generates antibody fluorescence distribution data; The detection information recording module receives the antibody fluorescence distribution data, analyzes and records the types, detection times and associated disease types of multiple antibodies in the image, and generates an antibody type analysis record.

[0044] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. The AI-based CBA fluorescent immunoassay method is characterized by: The following steps are involved: Based on the fluorescence immunoassay data, the image's light intensity and vibration errors are analyzed and identified, the exposure parameters and stabilization algorithm are adjusted to eliminate errors and noise, and an image correction processing record is generated; Based on the image correction processing record, edge information in the image is identified and enhanced, and the clarity of multiple key features in the image is optimized by combining contrast adjustment and sharpening processing to generate an image feature enhancement result; According to the image feature enhancement result, the cell and tissue features in the image are classified, and a plurality of pathological conditions are identified by analyzing the matching degree between the image features and the known pathological features, so as to generate a pathological condition classification result; According to the pathological state classification result, by analyzing the data features in the pathological image, matching pathological feature information for multiple pathological results, and generating a diagnostic information matching result; Based on the diagnostic information matching results, analyzing the fluorescence patterns of multiple antibodies in the feature-enhanced image, identifying the intensity, distribution and dynamic changes of the fluorescence of multiple antibodies, and obtaining fluorescence pattern analysis results; Based on the fluorescence pattern analysis results, through image recognition and biomarker database comparison, the antibody types present in the image are analyzed and the detection time and disease types of multiple antibodies are recorded, and the antibody type analysis record is output.

2. The AI-based CBA fluorescent immunoassay method according to claim 1, characterized in that: The image correction processing record includes the adjusted exposure parameters, the eliminated vibration errors, and the reduced number of noise points; the image feature enhancement results include enhanced edge information, adjusted contrast parameters, and sharpened key features; the pathological state classification results include classified cell types, identified tissue features, and matched pathological states; the diagnostic information matching results include corresponding health risk levels, etiology-related information, and referenced pathological data; the fluorescence pattern analysis results include identified antibody fluorescence intensity, calculated spatial distribution, and measured relative brightness; and the antibody type analysis record includes recorded antibody type, detection time, and associated pathological type.

3. The AI-based CBA fluorescent immunoassay method according to claim 1, characterized in that: Based on the fluorescence immunoassay data, the image light intensity and vibration errors are analyzed and identified, the exposure parameters and stabilization algorithm are adjusted to eliminate errors and noise, and the steps to generate image correction processing records are as follows: According to the fluorescence immunoassay image data, analyze the light intensity of the image, record the brightness values ​​of multiple areas, adjust the exposure parameters to optimize the lighting effect, and generate brightness adjustment records; Based on the brightness adjustment record, analyzing the image vibration error, adjusting the image stabilization parameters to eliminate the vibration caused by the device and optimize the image stability, and generating a stability optimization record; Based on the stability optimization record, filter strength and parameters are adjusted to optimize random noise and color distortion in the image, and an image correction processing record is generated.

4. The AI-based CBA fluorescent immunoassay method according to claim 1, characterized in that: Based on the image correction processing record, edge information in the image is identified and enhanced, contrast adjustment and sharpening are combined to optimize the clarity of multiple key features in the image, and the steps of generating image feature enhancement results are specifically as follows: Based on the image correction processing record, edge information in the image is identified, including analyzing edge clarity and continuity to obtain edge identification results; Based on the edge recognition result, adjusting the contrast of multiple regions in the image to optimize the distinguishability of key features and generate a contrast processing result; Based on the contrast processing result, the image sharpening parameters are adjusted to optimize the clarity of multiple key features in the image, thereby obtaining an image feature enhancement result.

5. The AI-based CBA fluorescent immunoassay method according to claim 1, characterized in that: According to the image feature enhancement result, the cell and tissue features in the image are classified, and a plurality of pathological conditions are identified by analyzing the matching degree between the image features and the known pathological features. The specific steps of generating the pathological condition classification result are as follows: Based on the image feature enhancement result, analyzing the image, identifying the structural features of cells and tissues in the image, including cell shape, size and color, and generating a cell structure analysis result; Based on the cell structure analysis result, the cell and tissue features in the image are compared with the pathological feature database and the matching degree is calculated to obtain the pathological feature matching result; Based on the pathological feature matching results, the cell and tissue features in the image are classified, and multiple pathological states are identified according to the matching degree to obtain a pathological state classification result.

6. The AI-based CBA fluorescent immunoassay method according to claim 1, characterized in that: Based on the diagnostic information matching result, the steps of analyzing the fluorescence patterns of multiple antibodies in the feature-enhanced image, identifying the intensity, distribution and dynamic changes of the fluorescence of multiple antibodies, and obtaining the fluorescence pattern analysis result are specifically as follows: Based on the diagnostic information matching result, performing spectral analysis on the fluorescence image, identifying the fluorescence intensity of multiple antibodies, recording the spectral data of the multiple antibodies, and generating fluorescence intensity data; Based on the fluorescence intensity data, a kernel density estimation algorithm is used to calculate the spatial distribution of antibody fluorescence in the image, evaluate the uniformity and aggregation area of ​​fluorescence distribution, and obtain fluorescence spatial distribution information; Based on the fluorescence spatial distribution information, the dynamic changes of fluorescence are analyzed, the fluorescence change trend in the time series is evaluated, and the fluorescence pattern analysis result is obtained.

7. The AI-based CBA fluorescent immunoassay method according to claim 6, characterized in that: The kernel density estimation algorithm is based on the formula: ； Calculate the fluorescence density estimate, where For the evaluation point The estimated density value at is the total number of data points, For the evaluation point, For the The location of the data points, is the global bandwidth, To dynamically adjust parameters, is the spatial resolution factor, Adjust the parameters for the kernel shape, is the kernel function.

8. The AI-based CBA fluorescent immunoassay method according to claim 1, characterized in that: Based on the fluorescence pattern analysis results, the antibody types present in the image are analyzed and the detection time and disease types of multiple antibodies are recorded through image recognition and biomarker database comparison. The specific steps of outputting the antibody type analysis record are as follows: Based on the fluorescence pattern analysis results, analyzing and identifying spectral features of multiple antibody fluorescence images, including intensity features and distribution patterns, and generating antibody spectral feature records; Based on the antibody spectral feature records, the spectral data of each antibody is analyzed using image recognition, and compared with the antibody features in a known biomarker database to obtain an antibody matching result; According to the antibody matching results, the detection time and associated disease types of the multiple antibodies are recorded to generate an antibody type analysis record.

9. The AI-based CBA fluorescent immune judgment system is characterized by: The AI-based CBA fluorescent immunoassay judgment system is used to execute the AI-based CBA fluorescent immunoassay judgment method according to any one of claims 1 to 6, and the AI-based CBA fluorescent immunoassay judgment system includes: The exposure parameter adjustment module collects fluorescent immunoassay image data, analyzes the light intensity and vibration errors of multiple areas in the image, adjusts the exposure parameters and eliminates image errors and noise, and generates image correction results; The edge information optimization module receives the image correction result, identifies the edge information in the image, and optimizes the clarity of key features in the image by combining contrast adjustment and sharpening processing to generate key feature processing results; The pathology recognition and classification module classifies cell and tissue features based on the key feature processing results, identifies multiple pathological states matching pathological feature information, including health risks and causes, by comparing the matching degree between image features and known pathological features, and generates pathological information diagnosis records; The fluorescence pattern evaluation module analyzes the fluorescence patterns of multiple antibodies in the feature-enhanced image based on the pathological information diagnosis record, identifies the intensity, distribution and dynamic changes of the fluorescence, calculates the spatial distribution and relative brightness of the fluorescence, and generates antibody fluorescence distribution data; The detection information recording module receives the antibody fluorescence distribution data, analyzes and records the types, detection times and associated disease types of multiple antibodies in the image, and generates an antibody type analysis record.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the AI-based CBA fluorescent immunoassay method according to any one of claims 1 to 6 are implemented.

Images