ANA detection image-text report intelligent generation method and system

Through the combination of multi-labeling and multi-wavelength fluorescence imaging technology, the sensitivity and consistency problems of traditional ANA detection are solved, and efficient and accurate antibody response pattern recognition and report generation are achieved.

CN120340736APending Publication Date: 2025-07-18GUANGZHOU HUISHAN MEDICAL TECH CO LTD +1
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202510292270.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-03-12
Publication Date
2025-07-18

AI Technical Summary

Technical Problem

Traditional ANA detection methods have high subjectivity, low sensitivity and inconsistency in detection results, especially when processing complex samples, they are prone to external interference, and lack of automation and intelligence.

Method used

The multi-labeling module uses high-sensitivity fluorescent markers such as quantum dots and nanoparticles, combined with multi-wavelength excitation light source and time-resolved fluorescence imaging technology, and uses a convolutional neural network to identify the antibody reaction mode and automatically generate detailed detection reports.

Benefits of technology

It significantly improves the sensitivity and accuracy of ANA detection, reduces artificial errors, and realizes intelligent analysis of antibody reactions and fast and accurate detection report generation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120340736A_ABST
    Figure CN120340736A_ABST
Patent Text Reader

Abstract

The invention discloses an ANA detection image-text report intelligent generation method and system, and relates to the technical field of immunodetection, the system uses different fluorescent markers to mark ANA, signal sensitivity is enhanced, and detection accuracy is ensured. Then, collecting fluorescence signals of different wavelengths by adopting a fluorescence microscope and a multi-wavelength excitation light source, and obtaining dynamic change data of reaction intensity by combining a time-resolved fluorescence imaging technology to generate a multi-channel imaging data set; thirdly, noise removal, signal enhancement and image alignment are carried out through an image preprocessing technology, and data consistency is ensured; based on a multi-channel data set, a convolutional neural network is used for training and testing an initial model, an intelligent detection model is constructed, a fluorescence reaction mode is extracted, finally, a detailed ANA detection report is automatically generated by calculating a comprehensive scoring coefficient and comparing the comprehensive scoring coefficient with a preset threshold value, and an accurate immunodiagnosis basis is provided for doctors.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of immunoassay technology, and specifically to an intelligent generation method and system for ANA detection graphic reports. Background Art

[0002] With the rapid development of medical technology and intelligent devices, the detection of antinuclear antibodies (ANA) has become an important tool for the diagnosis of autoimmune diseases. Traditional ANA detection methods generally rely on directly observing the fluorescence reaction pattern in the patient's serum sample under a microscope. However, limited by technical level and operating environment, traditional methods have high subjectivity and low detection sensitivity. This makes it difficult to ensure the accuracy and consistency of the results, and when dealing with complex samples, it is easily affected by external interference.

[0003] In recent years, with the introduction of high-sensitivity fluorescent markers such as quantum dots and nanoparticles, traditional fluorescence microscopy technology has been significantly improved. Through multiplex labeling and multi-wavelength fluorescence imaging technology, weak signal reactions can be enhanced and the accuracy and sensitivity of detection can be improved. At the same time, the progress of image acquisition and analysis technology has also made it possible to capture the reaction process in real time and monitor it dynamically, further improving the detection accuracy. However, there is still a large room for development in the automation degree and intelligent level of existing ANA detection systems, especially the deficiencies in automated report generation and risk assessment. Summary of the Invention

[0004] Aiming at the deficiencies of the prior art, the present invention provides an intelligent generation method and system for ANA detection graphic reports to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention is realized through the following technical solutions: an intelligent generation method and system for ANA detection graphic reports, including: a multiplex labeling module, an image acquisition module, a data processing module, an intelligent detection model module, and an intelligent report generation module; The multiplex labeling module is used to label the antinuclear antibody ANA in the sample with a variety of fluorescent markers, through quantum dots, nanoparticles or other high-sensitivity markers, to label different types of antibody reactions, so as to obtain corresponding fluorescence signals at different wavelengths, and enhance weak signal reactions to ensure detection sensitivity and accuracy; The image acquisition module is used to use a high-performance fluorescence microscope equipped with a multi-wavelength excitation light source to collect different types of antibody reaction patterns respectively, and combine time-resolved fluorescence imaging technology to obtain the change of reaction intensity at different time points, so as to improve the accuracy of reaction pattern recognition, capture the dynamic process of the reaction in real time, and generate a complete multi-channel imaging data set; The data processing module is used to preprocess the collected multi-channel image data, including noise removal, signal enhancement, and image alignment, to make the data consistent; The intelligent detection model module is used to train and test the initial convolutional neural network model based on the multi-channel image data set, and then calculate the antibody reaction intensity index KQD, the reaction area density index FYM, and the reaction area morphology index FXZ as the output of the middle layer as the feature vector, establish an intelligent detection model by identifying the feature information, and extract the fluorescence reaction patterns including: Homogeneous, Speckled, Peripheral, and Centromer; The intelligent report generation module is used to calculate the comprehensive scoring coefficient Szh, compare and analyze it with the preset threshold, and automatically generate a detailed ANA detection report according to the evaluation result. The report content includes: antibody reaction type, intensity and area, and the area and density of the reaction area.

[0006] Preferably, the multiple labeling module includes a light source signal acquisition unit, a light source signal calculation unit, and a light source signal analysis unit; The light source signal acquisition unit is used to measure the intensity value Sig of the fluorescence signal and the signal enhancement value Esen of the marker using a spectrometer, measure the power value Pin of the input light source using a power meter, and obtain the conversion efficiency Tsen of the image sensor using a quantum efficiency test; The light source signal calculation unit is used to monitor the influence of various markers on the fluorescence signal in real time. After dimensionless processing, calculate and obtain the fluorescence signal sensitivity coefficient Sfx, and the formula is as follows: ; In the formula, n represents the number of types of markers, Sig i represents the intensity value of the fluorescence signal of the i-th marker, Esen i represents the signal enhancement value of the i-th marker, Pin i represents the power value of the input light source of the i-th marker, Tsen i represents the conversion efficiency of the image sensor of the i-th marker, a i represents the absorption coefficient of the i-th marker, which is obtained through experiments. The common values are as follows in the table:

[0007] The light source signal analysis unit is used to preset the first standard threshold Q1, and compare and analyze the fluorescence signal sensitivity coefficient Sfx with the first standard threshold Q1 to obtain the first evaluation result including: When the fluorescence signal sensitivity coefficient Sfx < the first standard threshold Q1, it indicates that the signal emitted by the current marker cannot be clearly distinguished, triggering the first warning instruction, and the current marker cannot be used; When the fluorescence signal sensitivity coefficient Sfx ≥ the first standard threshold Q1, it indicates that the signal emitted by the current marker can be clearly distinguished, and a qualified label is attached to the marker.

[0008] Preferably, the image acquisition module includes an excitation light source setting unit, a fluorescence imaging unit, and a time-resolved fluorescence imaging unit; The excitation light source setting unit is used to select the markers with qualified labels and set different wavelength fluorescence markers to label antinuclear antibodies respectively; The fluorescence imaging unit is used to use a high-performance optical microscope equipped with a multi-wavelength excitation light source to prepare the labeled serum sample into a suitable thin slice, place it on the microscope slide, and use a multi-wavelength fluorescence microscope to simultaneously capture the fluorescence signals emitted by different markers. The signals are separated and recorded through a filter, generating multi-channel image data; The time-resolved fluorescence imaging unit is used to use a high-speed imaging device to capture the change curve of the reaction intensity at different time points, analyze the temporal characteristics of the antibody binding reaction, and obtain a multi-channel image data set.

[0009] Preferably, the time-resolved fluorescence imaging unit includes a fluorescence decay calculation sub-unit and a fluorescence decay analysis sub-unit; The fluorescence decay calculation sub-unit is used to continuously monitor the change in the captured reaction intensity, set to collect once every 5 minutes, combine the data of the light source signal acquisition unit, and calculate and obtain the fluorescence decay coefficient YGsj after dimensionless processing. The formula is as follows: ; In the formula, t represents time, Sig o represents the initial fluorescence signal intensity value, Sig t represents the fluorescence signal intensity value at time t, Pin o represents the power value of the initial input light source, Pin t represents the power value of the input light source at time t, D o represents the diffusion coefficient of the initial temperature antibody in the solution, D t represents the diffusion coefficient of the temperature antibody in the solution at time t, and w1, w2, and w3 represent weight coefficients; ; ; In the formula, kb represents the Boltzmann constant, and the common value is , Te o represents the initial temperature value, Te tRepresents the temperature value at time t, represents the viscosity value of the solution, and r represents the radius of the marker The fluorescence decay analysis subunit is used to preset a second standard threshold Q2, and compare and analyze the fluorescence decay coefficient YGsj with the second standard threshold Q2 to obtain a second evaluation result, including: When the fluorescence decay coefficient YGsj > the second standard threshold Q2, it indicates that the fluorescence signal decay accelerates, there is a risk of affecting data instability, and a second warning instruction is triggered. The first strategy generated includes: using an automatic constant temperature adjustment device to keep the temperature constant; reducing the initial fluorescence signal intensity value by 20%; reducing the power value of the initial input light source by 10%; increasing the image acquisition frequency by 30% to capture more time points of the reaction to avoid losing important information due to signal decay; When the fluorescence decay coefficient YGsj ≤ the second standard threshold Q2, it indicates that the fluorescence signal decay does not accelerate and there is no risk of affecting data instability, and a multi-channel image data set is generated.

[0010] Preferably, the data processing module includes an image processing unit; The data processing unit is used to preprocess the multi-channel image data set, remove environmental noise through Gaussian filtering, then perform signal enhancement using the histogram equalization method, and finally use image registration technology to align the images at different time points and wavelengths to ensure data consistency.

[0011] Preferably, the intelligent detection model module includes a fluorescence reaction mode calculation unit and a model establishment unit; The fluorescence reaction mode calculation unit is used to monitor the data of the multi-channel image data set and the data collected by the light source signal acquisition unit. After dimensionless processing, the antibody reaction intensity index KQD, the reaction area density index FYM, and the reaction area morphology index FXZ are calculated respectively. The formulas are as follows: ; In the formula, Esen represents the signal enhancement value of the marker, Pin represents the power value of the input light source of the marker, Tsen represents the conversion efficiency of the marker image sensor, a represents the absorption coefficient of the marker, and A represents the area of the reaction area; ;

[0012] In the formula, Sig represents the fluorescence signal intensity value; ;

[0013] In the formula, P is the perimeter of the reaction area, C is the roundness index, represents the radius of the reaction area.

[0014] Preferably, the model establishment unit is configured to use a convolutional neural network to construct an initial convolutional neural network model, train and test the initial convolutional neural network model with a multi-channel image data set, and use the trained initial convolutional neural network model as an intelligent detection model. At the same time, the intermediate layer outputs of the antibody reaction intensity index, reaction area density index, and reaction area morphology index are used as feature vectors to identify feature information, and the intelligent detection model is trained and tested with the obtained feature information, and the fluorescence reaction patterns extracted include: Homogeneous, Speckled, Peripheral, and Centromer.

[0015] Preferably, the intelligent report generation module includes a report data calculation unit and a report data analysis unit; The report data calculation unit is configured to calculate and obtain a comprehensive scoring coefficient Szh after dimensionless processing of the data extracted by the intelligent detection model. The formula is as follows: ; In the formula, DD represents the titer ratio, Sms represents the fluorescence pattern type score, and w4 and w5 represent the weight coefficients; ; In the formula, KT represents the antibody concentration value, and KT represents the antibody concentration value of the initial sample; ; In the formula, Homogeneous represents the homogeneous fluorescence reaction pattern, Speckled represents the speckled fluorescence reaction pattern, Peripheral represents the peripheral fluorescence reaction pattern, Centromer represents the centromeric fluorescence reaction pattern, and a1, a2, a3, and a4 respectively represent the scores for Homogeneous, Speckled, Peripheral, and Centromer.

[0016] Preferably, the report data analysis unit is configured to preset a third standard threshold Q3, and compare and analyze the comprehensive scoring coefficient Szh with the third standard threshold Q3 to obtain a third evaluation result, including: When the comprehensive scoring coefficient Szh < the third standard threshold Q3, it indicates that there is no risk of autoimmune disease, and continuous monitoring is required; When the comprehensive scoring coefficient Szh ≥ the third standard threshold Q3, it indicates that there is a risk of autoimmune disease, and a detailed ANA test report is automatically generated. The report content includes: antibody reaction type, intensity and area, area and density of the reaction area.

[0017] Preferably, an intelligent generation method for ANA detection graphic report includes the following steps: Step 1: Label the anti-nuclear antibody ANA in the sample with different fluorescent markers to enhance signal sensitivity and ensure the accuracy of detection; Step 2: Use a fluorescence microscope and a multi-wavelength excitation light source to collect fluorescence signals of different wavelengths respectively, and combine time-resolved fluorescence imaging technology to obtain the reaction intensity change data at different time points and generate a complete multi-channel imaging data set; Step 3: Preprocess the collected multi-channel image data, including noise removal, signal enhancement and image alignment, to make the data consistent; Step 4: Train and test the initial model of the convolutional neural network based on the multi-channel image data set, and then use the antibody reaction intensity index, reaction area density index, and reaction area morphology index as the output of the middle layer as the feature vector to establish an intelligent detection model by identifying feature information and extract the fluorescence reaction pattern; Step 5: Calculate the comprehensive scoring coefficient, and then compare and analyze it with the preset threshold. According to the evaluation result, automatically generate a detailed ANA detection report.

[0018] The present invention provides an intelligent generation method and system for ANA detection graphic report. It has the following beneficial effects: (1) The intelligent generation method and system for ANA detection graphic report significantly enhance the weak signal capture ability by using high-sensitivity fluorescent markers such as quantum dots and nanoparticles, combined with a multi-wavelength excitation light source and time-resolved fluorescence imaging technology, improve the sensitivity and accuracy of anti-nuclear antibody (ANA) detection, and ensure more reliable detection results.

[0019] (2) The intelligent generation method and system for ANA detection graphic report preprocess the multi-channel images through a data processing module, including noise removal, signal enhancement and image alignment, effectively improve the data consistency and image quality, reduce human errors, and ensure the accuracy of subsequent analysis.

[0020] (3) The intelligent generation method and system for ANA detection graphic report, based on the intelligent detection model of the convolutional neural network, can accurately identify different antibody reaction patterns, automatically extract features such as reaction intensity, area density and morphology, realize the intelligent analysis of antibody reactions, reduce human operation errors, and improve the detection efficiency and accuracy.

[0021] (4)The intelligent generation method and system for ANA detection graphic reports can automatically generate detailed ANA detection reports by comparing and analyzing the comprehensive scoring coefficient with the preset threshold. The reports include information such as antibody reaction type, intensity, area, and density of the reaction region, providing accurate and rapid diagnostic support for clinical practice and improving the reliability and real-time nature of the detection reports. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a schematic block diagram flow chart of an intelligent generation system for ANA detection graphic reports of the present invention; Figure 2 It is a schematic diagram of the steps of an intelligent generation method for ANA detection graphic reports of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0023] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0024] Embodiment 1 Please refer to Figure 1 , the present invention provides an intelligent generation system for ANA detection graphic reports, including: a multiple labeling module, an image acquisition module, a data processing module, an intelligent detection model module, and an intelligent report generation module; The multiple labeling module is used to label the antinuclear antibody ANA in the sample with multiple fluorescent markers. Through quantum dots, nanoparticles, or other highly sensitive markers, different types of antibody reactions are labeled, so as to obtain corresponding fluorescence signals at different wavelengths and enhance weak signal reactions to ensure detection sensitivity and accuracy; The image acquisition module is used to use a high-performance fluorescence microscope equipped with a multi-wavelength excitation light source to collect antibody reaction patterns of different types respectively, and combine time-resolved fluorescence imaging technology to obtain the change of reaction intensity at different time points, so as to improve the accuracy of reaction pattern recognition, capture the dynamic process of the reaction in real time, and generate a complete multi-channel imaging data set; The data processing module is used to preprocess the collected multi-channel image data, including noise removal, signal enhancement, and image alignment, to make the data consistent; The intelligent detection model module is used to train and test the initial convolutional neural network model based on a multi-channel image dataset. Then, the antibody reaction intensity index KQD, reaction area density index FYM, and reaction area morphology index FXZ are calculated and used as the output of the middle layer as feature vectors to establish an intelligent detection model by identifying feature information. The extracted fluorescence reaction patterns include: Homogeneous, Speckled, Peripheral, and Centromer; The intelligent report generation module is used to calculate and obtain the comprehensive scoring coefficient Szh, compare it with a preset threshold for analysis, and automatically generate a detailed ANA detection report based on the evaluation results. The report content includes: antibody reaction type, intensity, and area, as well as the area and density of the reaction area.

[0025] In this embodiment, by using the multiple labeling module to perform multiple fluorescence labelings on the antinuclear antibody ANA in the sample, the sensitivity and accuracy of the fluorescence signal are improved. Especially in the case of weak signals, the reaction signal can be effectively enhanced, thus ensuring high sensitivity and high accuracy of the detection, which helps to improve the recognition accuracy of the antibody reaction and the reliability of the detection results. The intelligent detection model based on the convolutional neural network can accurately identify different antibody reaction patterns, automatically extract features such as reaction intensity, area density, and morphology, realizing the intelligent analysis of the antibody reaction, reducing human operation errors, and improving the detection efficiency and accuracy. Through the comparison and analysis of the comprehensive scoring coefficient with the preset threshold, the system can automatically generate a detailed ANA detection report, providing accurate and rapid diagnostic support for clinical practice, and improving the reliability and real-time nature of the detection report.

[0026] Example 2. This example is an explanatory description based on Example 1. Please refer to Figure 1 , specifically, the multiple labeling module includes a light source signal acquisition unit, a light source signal calculation unit, and a light source signal analysis unit; The light source signal acquisition unit is used to measure the intensity value Sig of the fluorescence signal and the signal enhancement value Esen of the marker using a spectrometer, measure the power value Pin of the input light source using an optical power meter, and obtain the conversion efficiency Tsen of the image sensor using quantum efficiency testing; The light source signal calculation unit is used to monitor the influence of various markers on the fluorescence signal in real time. After dimensionless processing, the fluorescence signal sensitivity coefficient Sfx is calculated as follows: ;

[0027] In the formula, n represents the number of types of markers, Sig i represents the intensity value of the fluorescence signal of the i-th marker, Esen iThe signal enhancement value of the i-th marker, Pin i The power value of the input light source of the i-th marker, Tsen i The conversion efficiency of the image sensor of the i-th marker, a i The absorption coefficient of the i-th marker, obtained through experiments, and the common values are as follows in the table:

[0028] The light source signal analysis unit is used to preset a first standard threshold Q1, and compare and analyze the fluorescence signal sensitivity coefficient Sfx with the first standard threshold Q1 to obtain the first evaluation result, including: When the fluorescence signal sensitivity coefficient Sfx < the first standard threshold Q1, it indicates that the signal emitted by the current marker cannot be clearly distinguished, triggering a first warning instruction, and the current marker cannot be used; When the fluorescence signal sensitivity coefficient Sfx ≥ the first standard threshold Q1, it indicates that the signal emitted by the current marker can be clearly distinguished, and a qualified label is attached to the marker.

[0029] In this embodiment, through the collaborative work of the light source signal acquisition, calculation, and analysis unit, the fluorescence signal intensity and sensitivity of the marker can be monitored and evaluated in real time, ensuring that only those markers that can be clearly distinguished are used, avoiding misjudgment or inaccurate detection caused by unclear signals, and improving the accuracy and reliability of the detection results.

[0030] Embodiment 3. This embodiment is an explanatory description carried out in Embodiment 2. Please refer to Figure 1 , specifically, the image acquisition module includes an excitation light source setting unit, a fluorescence imaging unit, and a time-resolved fluorescence imaging unit; The excitation light source setting unit is used to select the markers with qualified labels and set different wavelength fluorescence markers to label anti-nuclear antibodies respectively; The fluorescence imaging unit is used to use a high-performance fluorescence microscope equipped with a multi-wavelength excitation light source, prepare the labeled serum sample into a suitable thin slice, place it on the microscope slide, and use the multi-wavelength fluorescence microscope to simultaneously capture the fluorescence signals emitted by different markers. The signals are separated and recorded through a filter to generate multi-channel image data; The time-resolved fluorescence imaging unit is used to use a high-speed imaging device to capture the change curve of the reaction intensity at different time points and analyze the temporal characteristics of the antibody binding reaction to obtain a multi-channel image data set.

[0031] In this embodiment, through the collaborative work of the excitation light source setting unit, the fluorescence imaging unit, and the time-resolved fluorescence imaging unit, the fluorescence signal of antinuclear antibody can be accurately labeled and captured at different wavelengths, and the temporal change of the reaction intensity can be recorded in real time, so as to generate a high-quality multi-channel image dataset. This not only improves the accuracy of reaction pattern recognition, but also can accurately capture the dynamic process of antibody binding reaction, which helps to comprehensively and deeply analyze the characteristics of antibody reaction.

[0032] Example 4. This example is an explanatory description based on Example 3. Please refer to Figure 1 , specifically, the time-resolved fluorescence imaging unit includes a fluorescence decay calculation sub-unit and a fluorescence decay analysis sub-unit; The fluorescence decay calculation sub-unit is used to continuously monitor the change of the captured reaction intensity, set to collect once every 5 minutes, combine the data of the light source signal acquisition unit, and calculate the fluorescence decay coefficient YGsj after dimensionless processing. The formula is as follows: ; In the formula, t represents time, Sig o represents the initial fluorescence signal intensity value, Sig t represents the fluorescence signal intensity value at time t, Pin o represents the power value of the initial input light source, Pin t represents the power value of the input light source at time t, D o represents the diffusion coefficient of the initial temperature antibody in the solution, D t represents the diffusion coefficient of the temperature antibody in the solution at time t, w1, w2, and w3 represent weight coefficients, , , , and ; ; ; In the formula, kb represents the Boltzmann constant, and the common value is , Te o represents the initial temperature value, Te t represents the temperature value at time t, represents the viscosity value of the solution, and r represents the radius of the marker; The fluorescence decay analysis sub-unit is used to preset the second standard threshold Q2, and compare and analyze the fluorescence decay coefficient YGsj with the second standard threshold Q2 to obtain the second evaluation result, including: When the fluorescence decay coefficient YGsj > the second standard threshold Q2, it indicates that the fluorescence signal decay accelerates, there is a risk of affecting data instability, and the second warning instruction is triggered. The first strategy generated includes: using an automatic constant temperature adjustment device to keep the temperature constant; reducing the initial fluorescence signal intensity value by 20%; reducing the power value of the initial input light source by 10%; increasing the image acquisition frequency by 30% to capture more time points of the reaction to avoid losing important information due to signal decay; When the fluorescence decay coefficient YGsj ≤ the second standard threshold Q2, it indicates that the fluorescence signal decay does not accelerate and there is no risk of affecting data instability, and a multi-channel image data set is generated.

[0033] In this embodiment, through the real-time monitoring and calculation of the fluorescence decay calculation subunit and the fluorescence decay analysis subunit, the decay of the fluorescence signal during the reaction process can be accurately tracked, and the warning mechanism can be automatically triggered according to the set standard threshold, and the experimental parameters can be adjusted in a timely manner, such as the optimization of constant temperature adjustment, signal intensity, and acquisition frequency, to ensure the stability of the fluorescence signal and maximize the capture of important reaction information. In this way, the system can effectively avoid data instability caused by too fast decay of the fluorescence signal, thereby ensuring the accuracy and reliability of the experimental results.

[0034] Example 5. This example is an explanatory description carried out in Example 4. Please refer to Figure 1 , specifically, the data processing module includes an image processing unit; The data processing unit is used to perform image preprocessing on the multi-channel image data set, remove environmental noise through Gaussian filtering, then perform signal enhancement using the histogram equalization method, and finally use image registration technology to align the images at different time points and wavelengths to ensure data consistency.

[0035] In this embodiment, through the Gaussian filtering, histogram equalization, and image registration technology of the image processing unit, environmental noise can be effectively removed and the signal can be enhanced, while ensuring the alignment of image data at different time points and wavelengths, thereby ensuring the high consistency and high quality of the multi-channel image data set. This image preprocessing method improves the accuracy of subsequent analysis and enhances the reliability of the detection results.

[0036] Example 6. This example is an explanatory description carried out in Example 5. Please refer to Figure 1 , specifically, the intelligent detection model module includes a fluorescence reaction mode calculation unit and a model establishment unit; The fluorescence reaction mode calculation unit is used to monitor the data of the multi-channel image data set and the data collected by the light source signal acquisition unit. After dimensionless processing, the antibody reaction intensity index KQD, the reaction area density index FYM, and the reaction area morphology index FXZ are calculated respectively. The formulas are as follows: ;

[0037] In the formula, Esen represents the signal enhancement value of the marker, Pin represents the power value of the input light source of the marker, Tsen represents the conversion efficiency of the marker image sensor, a represents the absorption coefficient of the marker, and A represents the area of the reaction region; ; In the formula, Sig represents the fluorescence signal intensity value; ;

[0038] In the formula, P is the perimeter of the reaction region, C is the roundness index, represents the radius of the reaction region.

[0039] In this embodiment, through the accurate calculation of the antibody reaction intensity index KQD, the reaction region density index FYM, and the reaction region morphology index FXZ by the fluorescence reaction mode calculation unit, the reaction characteristics of antinuclear antibodies in different markers and reaction regions can be quantified and evaluated. This data analysis method after dimensionless processing provides a more refined reaction mode recognition ability, helps to improve the accuracy and sensitivity of the intelligent detection model, and ensures more reliable antibody reaction evaluation results.

[0040] Example 7. This example is an explanatory description carried out in Example 6. Please refer to Figure 1 , specifically, the model establishment unit is used to construct an initial convolutional neural network model using a convolutional neural network, train and test the initial convolutional neural network model with a multi-channel image data set, and use the trained initial convolutional neural network model as an intelligent detection model. At the same time, the intermediate layer outputs of the antibody reaction intensity index, the reaction region density index, and the reaction region morphology index are used as feature vectors to identify feature information, train and test the intelligent detection model with the obtained feature information, and extract fluorescence reaction modes including: Homogeneous, Speckled, Peripheral, and Centromer.

[0041] In this embodiment, through the training and testing of the initial convolutional neural network model and the combination of the antibody reaction intensity index, the reaction region density index, and the reaction region morphology index as feature vectors, the recognition accuracy and generalization ability of the intelligent detection model can be effectively improved. This method can not only accurately extract different types of fluorescence reaction modes including Homogeneous, Speckled, Peripheral, and Centromer, but also automatically identify the reaction mode of antinuclear antibodies, greatly improving the accuracy and efficiency of detection, and providing stronger intelligent support for antibody detection.

[0042] Example 8. This example is an explanatory note based on Example 7. Please refer to Figure 1 , the intelligent report generation module includes a report data calculation unit and a report data analysis unit; The report data calculation unit is used to calculate and obtain the comprehensive scoring coefficient Szh after dimensionless processing of the data extracted by the intelligent detection model. The formula is as follows: ; In the formula, DD represents the titer ratio, Sms represents the fluorescence mode type score, w4 and w5 represent the weight coefficients, , , and ; ; In the formula, KT represents the antibody concentration value, and KT o represents the antibody concentration value of the initial sample; ; In the formula, Homogeneous represents a homogeneous fluorescence reaction mode, Speckled represents a speckled fluorescence reaction mode, Peripheral represents a peripheral fluorescence reaction mode, Centromer represents a centromeric fluorescence reaction mode, and a1, a2, a3, and a4 respectively represent the scores for the homogeneous Homogeneous, speckled Speckled, peripheral Peripheral, and centromeric Centromer, , , , , and .

[0043] In this example, by combining the data extracted by the intelligent detection model with the calculation of the comprehensive scoring coefficient Szh, factors such as the titer ratio, fluorescence mode type score, and antibody concentration value can be comprehensively considered to automatically generate an accurate antibody detection report. This method can not only effectively evaluate the performance of different fluorescence reaction modes but also automatically adjust the report content according to the scoring results, providing a more comprehensive and accurate analysis basis for testers and significantly improving the standardization and automation of antibody detection.

[0044] Example 9. This example is an explanatory note based on Example 8. Please refer to Figure 1 , the report data analysis unit is used to preset the third standard threshold Q3 and compare and analyze the comprehensive scoring coefficient Szh with the third standard threshold Q3 to obtain the third evaluation result, including: When the comprehensive scoring coefficient Szh < the third standard threshold Q3, it indicates no risk of autoimmune disease, and continuous monitoring is required; When the comprehensive scoring coefficient Szh ≥ the third standard threshold Q3, it indicates a risk of autoimmune disease, and a detailed ANA test report is automatically generated. The report content includes: antibody reaction type, intensity, and area, as well as the area and density of the reaction area.

[0045] In this embodiment, through the comparative analysis of the comprehensive scoring coefficient Szh and the third standard threshold Q3, the system can automatically identify whether there is a risk of autoimmune disease. If the comprehensive scoring coefficient is lower than the preset threshold, the system will prompt continuous monitoring to avoid overdiagnosis; if the score is higher than the threshold, the system will automatically generate a detailed ANA test report, accurately listing key information such as antibody reaction type, intensity, and area, helping doctors quickly diagnose potential immune disease risks, thereby improving the diagnostic efficiency and accuracy.

[0046] Example 10. An intelligent generation method for ANA test graphic reports, please refer to Figure 2 , including the following steps: Step 1: Label the antinuclear antibody ANA in the sample with different fluorescent markers to enhance signal sensitivity and ensure the accuracy of detection; Step 2: Use a fluorescence microscope and a multi-wavelength excitation light source to collect fluorescence signals of different wavelengths respectively, and combine time-resolved fluorescence imaging technology to obtain reaction intensity change data at different time points, and generate a complete multi-channel imaging data set; Step 3: Preprocess the collected multi-channel image data, including noise removal, signal enhancement, and image alignment, to make the data consistent; Step 4: Based on the multi-channel image data set, train and test the initial model of the convolutional neural network, and then use the antibody reaction intensity index, reaction area density index, and reaction area morphology index as the intermediate layer output as the feature vector to establish an intelligent detection model by identifying feature information and extract the fluorescence reaction pattern; Step 5: Calculate the comprehensive scoring coefficient, and then conduct a comparative analysis with the preset threshold. According to the evaluation results, automatically generate a detailed ANA test report.

[0047] In this embodiment, through technical means such as multiple labeling, time-resolved fluorescence imaging, and intelligent model analysis, the method can efficiently and accurately detect the antinuclear antibody ANA and generate an accurate test report. By combining the convolutional neural network and reaction pattern analysis, the system can automatically identify the type, intensity, and regional characteristics of antibody reactions, providing a scientific scoring and diagnostic basis. At the same time, the automatically generated ANA test report provides doctors with detailed data support, helping to quickly and accurately diagnose potential immune diseases and significantly improving the detection efficiency and accuracy.

[0048] The setting of the threshold value is for the convenience of comparison. Regarding the size of the threshold value, it depends on the amount of sample data and the number of base quantities set by those skilled in the art for each group of sample data; as long as the proportional relationship between the parameter and the quantized value is not affected.

[0049] The above formulas are all obtained through software simulation by collecting a large amount of data and selecting a formula close to the true value. The coefficients in the formulas are set by those skilled in the art according to the actual situation. As mentioned above, the above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. An intelligent generation system for ANA detection graphic reports, characterized in that, Including: A multiple labeling module, an image acquisition module, a data processing module, an intelligent detection model module, and an intelligent report generation module; The multiple labeling module is used to label anti-nuclear antibody ANA in a sample using multiple fluorescent markers. Through quantum dots, nanoparticles, or other highly sensitive markers, different types of antibody reactions are labeled, so as to obtain corresponding fluorescence signals at different wavelengths, and enhance weak signal reactions to ensure detection sensitivity and accuracy; The image acquisition module is used to use a high-performance fluorescence microscope equipped with a multi-wavelength excitation light source to collect antibody reaction patterns of different types respectively, and combine time-resolved fluorescence imaging technology to obtain the change of reaction intensity at different time points, so as to improve the accuracy of reaction pattern recognition, capture the dynamic process of the reaction in real time, and generate a complete multi-channel imaging data set; The data processing module is used to preprocess the collected multi-channel image data, including noise removal, signal enhancement, and image alignment, to make the data consistent; The intelligent detection model module is used to train and test the initial convolutional neural network model based on the multi-channel image data set, and then calculate to obtain the antibody reaction intensity index KQD, the reaction area density index FYM, and the reaction area morphology index FXZ as the intermediate layer output as the feature vector, to establish an intelligent detection model by identifying feature information, and extract fluorescence reaction patterns including: Homogeneous, Speckled, Peripheral, and Centromer; The intelligent report generation module is used to calculate and obtain the comprehensive scoring coefficient Szh, and compare and analyze it with a preset threshold. According to the evaluation result, it automatically generates a detailed ANA detection report, and the report content includes: antibody reaction type, intensity and area, the area and density of the reaction area.

2. The intelligent generation system for ANA detection graphic report according to claim 1, characterized in that: The multiple labeling module includes a light source signal acquisition unit, a light source signal calculation unit, and a light source signal analysis unit; The light source signal acquisition unit is used to measure the intensity value Sig of the fluorescence signal and the signal enhancement value Esen of the marker using a spectrometer, measure the power value Pin of the input light source using a power meter, and obtain the conversion efficiency Tsen of the image sensor using a quantum efficiency test; The light source signal calculation unit is used to monitor the influence of various markers on the fluorescence signal in real time. After dimensionless processing, it calculates and obtains the fluorescence signal sensitivity coefficient Sfx, and the formula is as follows: ; Wherein, n represents the number of types of markers, Sig i represents the intensity value of the fluorescence signal of the i-th type of marker, Esen i represents the signal enhancement value of the i-th type of marker, Pin i represents the power value of the input light source of the i-th type of marker, Tsen i represents the conversion efficiency of the image sensor of the i-th type of marker, a i represents the absorption coefficient of the i-th type of marker, which is obtained through experiments; The light source signal analysis unit is used to preset a first standard threshold Q1, and compare and analyze the fluorescence signal sensitivity coefficient Sfx with the first standard threshold Q1 to obtain the first evaluation result including: When the fluorescence signal sensitivity coefficient Sfx < the first standard threshold Q1, it means that the signal emitted by the current marker cannot be clearly distinguished, triggers a first warning instruction, and the current marker cannot be used; When the fluorescence signal sensitivity coefficient Sfx ≥ the first standard threshold Q1, it means that the signal emitted by the current marker can be clearly distinguished, and a qualified label is attached to the marker.

3. An intelligent generation system for ANA detection graphic reports according to claim 2, characterized in that: The image acquisition module includes an excitation light source setting unit, a fluorescence imaging unit, and a time-resolved fluorescence imaging unit; The excitation light source setting unit is used to select the markers with qualified labels and set fluorescent markers with different wavelengths to label antinuclear antibodies respectively. The fluorescence imaging unit is used to prepare the labeled serum sample into a suitable thin slice with a high-performance fluorescence microscope equipped with a multi-wavelength excitation light source, place it on the microscope slide, and use the multi-wavelength fluorescence microscope to simultaneously capture the fluorescence signals emitted by different markers. The signals are separated and recorded through a filter to generate multi-channel image data. The time-resolved fluorescence imaging unit is used to use a high-speed imaging device to capture the change curve of the reaction intensity at different time points, analyze the temporal characteristics of the antibody binding reaction, and obtain a multi-channel image data set.

4. An intelligent generation system for ANA detection graphic reports according to claim 3, characterized in that: The time-resolved fluorescence imaging unit includes a fluorescence decay calculation sub-unit and a fluorescence decay analysis sub-unit. The fluorescence decay calculation sub-unit is used to continuously monitor the change of the capture reaction intensity, set to collect once every 5 minutes, combine the data of the light source signal acquisition unit, and after dimensionless processing, calculate and obtain the fluorescence decay coefficient YGsj. The formula is as follows: ; where t represents time, Sig o represents the initial fluorescence signal intensity value, Sig t represents the fluorescence signal intensity value at time t, Pin o represents the power value of the initial input light source, Pin t represents the power value of the input light source at time t, D o represents the diffusion coefficient of the initial temperature antibody in the solution, D t represents the diffusion coefficient of the temperature antibody in the solution at time t, and w1, w2, and w3 represent weighting coefficients; ; ; wherein, kb represents the Boltzmann constant, and the common value is , Te o represents the initial temperature value, Te t represents the temperature value at time t, represents the viscosity value of the solution, and r represents the radius of the marker; The fluorescence decay analysis sub-unit is used to preset the second standard threshold Q2 and compare and analyze the fluorescence decay coefficient YGsj with the second standard threshold Q2 to obtain the second evaluation result, including: When the fluorescence decay coefficient YGsj > the second standard threshold Q2, it indicates that the fluorescence signal decay accelerates, there is a risk of affecting data instability, trigger the second warning instruction, and generate the first strategy, including: using an automatic constant temperature adjustment device to keep the temperature unchanged; reducing the initial fluorescence signal intensity value by 20%; reducing the power value of the initial input light source by 10%; increasing the image acquisition frequency by 30% to capture more time points of the reaction to avoid losing important information due to signal decay. When the fluorescence decay coefficient YGsj ≤ the second standard threshold Q2, it indicates that the fluorescence signal decay does not accelerate and there is no risk of affecting data instability, and a multi-channel image data set is generated.

5. An intelligent generation system for ANA detection graphic reports according to claim 4, characterized in that: The data processing module includes an image processing unit. The data processing unit is used to preprocess the multi-channel image data set, remove environmental noise through Gaussian filtering, then perform signal enhancement using the histogram equalization method, and finally use image registration technology to align the images at different time points and wavelengths to ensure data consistency.

6. The intelligent generation system for ANA detection graphic report according to claim 5, characterized in that: The intelligent detection model module includes a fluorescence reaction mode calculation unit and a model establishment unit. The fluorescence reaction mode calculation unit is used to monitor the data of the multi-channel image data set and the data collected by the light source signal acquisition unit. After dimensionless processing, calculate and obtain the antibody reaction intensity index KQD, the reaction area density index FYM, and the reaction area morphology index FXZ respectively. The formula is as follows: ; In the formula, Esen represents the signal enhancement value of the marker, Pin represents the power value of the input light source of the marker, Tsen represents the conversion efficiency of the marker image sensor, a represents the absorption coefficient of the marker, and A represents the area of the reaction area. ; In the formula, Sig represents the fluorescence signal intensity value. ; Wherein, P is the perimeter of the reaction area, and C is the roundness index, representing the radius of the reaction area.

7. An intelligent generation system for ANA detection graphic reports according to claim 6, characterized in that: The model building unit is used to construct an initial convolutional neural network model using a convolutional neural network, train and test the initial convolutional neural network model with a multi-channel image dataset, and use the initial convolutional neural network model after training as an intelligent detection model. At the same time, the intermediate layer outputs of the antibody reaction intensity index, reaction area density index, and reaction area morphology index are used as feature vectors to identify feature information, and the intelligent detection model is trained and tested with the obtained feature information, and the fluorescence reaction patterns are extracted, including: Homogeneous, Speckled, Peripheral, and Centromer.

8. An intelligent generation system for ANA detection graphic reports according to claim 7, characterized in that: The intelligent report generation module includes a report data calculation unit and a report data analysis unit; The report data calculation unit is used to calculate and obtain a comprehensive scoring coefficient Szh after dimensionless processing of the data extracted by the intelligent detection model. The formula is as follows: ; In the formula, DD represents the titer ratio, Sms represents the fluorescence pattern type score, and w4 and w5 represent the weight coefficients; ; Where KT represents the antibody concentration value, KT o represents the antibody concentration value of the initial sample; ; In the formula, Homogeneous represents the homogeneous fluorescence reaction pattern, Speckled represents the speckled fluorescence reaction pattern, Peripheral represents the peripheral fluorescence reaction pattern, Centromer represents the centromeric fluorescence reaction pattern, and a1, a2, a3, and a4 respectively represent the scores for Homogeneous, Speckled, Peripheral, and Centromer.

9. An ANA detection graphic report intelligent generation system according to claim 8, characterized in that: The report data analysis unit is used to preset a third standard threshold Q3, and compare and analyze the comprehensive scoring coefficient Szh with the third standard threshold Q3 to obtain the third evaluation result, including: When the comprehensive scoring coefficient Szh < the third standard threshold Q3, it means there is no risk of autoimmune disease, and continuous monitoring is required; When the comprehensive scoring coefficient Szh ≥ the third standard threshold Q3, it means there is a risk of autoimmune disease, and a detailed ANA detection report is automatically generated. The report content includes: antibody reaction type, intensity and area, area and density of the reaction area.

10. A method for intelligently generating an ANA detection graphic report, including the ANA detection graphic report intelligent generation system according to any one of claims 1 to 9 above, characterized in that: It includes the following steps: Step 1: Use different fluorescent markers to label the antinuclear antibody ANA in the sample to enhance signal sensitivity and ensure the accuracy of detection; Step 2: Use a fluorescence microscope and a multi-wavelength excitation light source to collect fluorescence signals of different wavelengths respectively, and combine time-resolved fluorescence imaging technology to obtain reaction intensity change data at different time points to generate a complete multi-channel imaging dataset; Step 3: Preprocess the collected multi-channel image data, including noise removal, signal enhancement, and image alignment to make the data consistent; Step 4: Based on the multi-channel image dataset, train and test the initial convolutional neural network model, and then use the antibody reaction intensity index, reaction area density index, and reaction area morphology index as the intermediate layer output as the feature vector to identify feature information and establish an intelligent detection model, and extract the fluorescence reaction pattern; Step 5: Obtain the comprehensive scoring coefficient through calculation, and then conduct a comparative analysis with the preset threshold. According to the evaluation results, automatically generate a detailed ANA test report.

Citation Information

Cited By

  • Indirect immunofluorescence detection method and system for single person

    CN121027506A

  • Time-resolved fluorescence intensity analysis method, medium, system and application

    CN121705802A