Immunofluorescence image analysis method and related equipment

Abstract

The present invention provides an analysis method and related equipment for immunofluorescence images, which are applied to the field of data processing technology. The present application obtains target biological sample data and a training sample set matching the target biological sample data; processes the target biological sample data to generate a fluorescent staining image; obtains a preset fluorescent substance recognition model matching the fluorescent staining image; preprocesses the training sample set to generate a training sample set with target feature data; processes the preset fluorescent substance recognition model based on the training sample set with target feature data to generate a target fluorescent substance recognition model; processes the fluorescent staining image based on the target fluorescent substance recognition model to generate original partition image information; processes the original partition image information to generate fluorescent signal information; processes the fluorescent signal distribution area, fluorescent signal intensity and discreteness of the fluorescent signal to generate fluorescent image spatial distribution characteristics.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to an analysis method for immunofluorescence images and related equipment. Background Art

[0002] In the process of biological or medical research experiments, when it is necessary to verify the location and expression of a certain protein molecule or RNA molecule, an immunofluorescence experiment (for protein molecules) or a fluorescent in situ hybridization experiment (for RNA molecules) is required. This technology can be applied to research in various fields such as reproduction, tumors, cardiovascular, and neural development. By using antibodies to the target protein or fluorescent probes to the target RNA, the samples in the slices are marked, and then fluorescent images are obtained by taking them under a fluorescent microscope. Finally, the fluorescent images are analyzed to obtain the location and quantitative information of the target protein molecule or RNA molecule that the applicant wants.

[0003] In addition, when performing image analysis, it is necessary to pay attention to the distribution characteristics of the fluorescent signal. For example, in the field of reproduction, when it is necessary to count the coverage of XIST molecules on chromosomes, it is necessary to pay attention to the distribution state of the fluorescent signal, whether it is a clustered distribution or a diffuse distribution, which plays an important role in judging its function. However, there is currently a lack of methods on the market that can count the distribution characteristics and discreteness of fluorescent signals.

[0004] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0005] The purpose of the present application is to provide an analysis method for immunofluorescence images and related equipment and systems, which at least to a certain extent overcome the problems existing in the prior art, perform statistical analysis of quantitative characteristic values ​​for a specific area circled in a single-channel fluorescence image, provide various characteristic values ​​such as the area of ​​complex fluorescence signal distribution under different thresholds, the minimum circle area containing all fluorescence signals, the minimum circle radius and the coordinates of the center of the circle, and the discreteness of the fluorescence signal distribution, and further provide quantitative clues for the description and classification of the complex fluorescence signal distribution.

[0006] Other features and advantages of the present application will become apparent from the following detailed description, or may be learned in part by the practice of the present invention.

[0007] According to one aspect of the present application, a method for analyzing an immunofluorescence image is provided, comprising: acquiring target biological sample data and a training sample set matching the target biological sample data; processing the target biological sample data to generate a fluorescent staining image, wherein the fluorescent staining image is an image of a single fluorescent channel; acquiring a preset fluorescent substance recognition model matching the fluorescent staining image; preprocessing the training sample set to generate a training sample set with target feature data; processing the preset fluorescent substance recognition model based on the training sample set with target feature data to generate a target fluorescent substance recognition model; processing the fluorescent staining image based on the target fluorescent substance recognition model to generate original partition image information, wherein the original partition image information includes an original partition contour image and a fluorescent signal image; processing the original partition image information to generate fluorescent signal information, wherein the fluorescent signal information includes a fluorescent signal distribution area, a fluorescent signal intensity, and a discreteness of the fluorescent signal; processing the fluorescent signal distribution area, the fluorescent signal intensity, and the discreteness of the fluorescent signal to generate a fluorescent image spatial distribution feature.

[0008] In one embodiment of the present application, the training sample set is preprocessed to generate a training sample set with target feature data, including: extracting features from the training sample set to determine an original feature library; dividing each feature data set according to the original feature library to generate a training data set and a verification data set; using a classifier to predict each verification data set divided from the original feature library to determine a prediction result; using a preset algorithm to train each training data set divided from the original feature library to obtain a verification set class prediction result; generating target feature data based on the prediction result and the verification set class prediction result, wherein the target feature data is used to characterize the spatial distribution characteristics of the fluorescence image.

[0009] In one embodiment of the present application, the preset fluorescent substance recognition model is processed based on the training sample set with target feature data to generate a target fluorescent substance recognition model, including: dividing the training sample set with target feature data into a training set for training the preset fluorescent substance recognition model and a verification set for verifying the preset fluorescent substance recognition model based on a preset ratio; extracting multiple data groups from the training set, wherein each data group contains a preset number of data samples, wherein at least one data sample includes target feature data; training the preset fluorescent substance recognition model based on the data samples in the multiple data groups to generate a trained fluorescent substance recognition model; processing the trained fluorescent substance recognition model based on the verification set to generate a verification result; if the data samples containing the target feature data in the verification result are fluorescent pixels and the brightness of the fluorescent pixels, then the trained fluorescent substance recognition model is used as the target fluorescent substance recognition model.

[0010] In one embodiment of the present application, multiple groups of data groups are extracted from the training set, including: dividing the training set based on target feature data to generate a number of number of class samples; obtaining any number of class samples in the training set that are less than a preset threshold; generating adjacent samples based on the distance between any number of class samples that are less than the preset threshold and other number of class samples of the same category that are less than the preset threshold, wherein the adjacent samples include a preset number of any number of class samples that are less than the preset threshold; determining a sampling ratio based on the number of samples of each class in the training set; determining a sampling ratio based on the sampling ratio; sampling the adjacent samples based on the sampling ratio to generate a preset number of sampled samples; generating multiple groups of data groups based on any class of samples and each sampled sample.

[0011] In one embodiment of the present application, the original partition image information is processed to generate fluorescence signal information, including: preprocessing the original partition image information to generate partition image information in a target format, wherein the partition image information in the target format includes a partition contour image and a binary image of the fluorescence signal, and the binary image of the fluorescence signal includes fluorescent pixels and the sum of the brightness of the fluorescent pixels; graying the partition contour image to generate a grayscale image; binarizing the grayscale image to generate the coordinates of the outermost contour endpoints; generating partition contour area information based on the outermost contour endpoint coordinates; processing the partition contour area information based on preset processing rules to generate fluorescence signal information.

[0012] In one embodiment of the present application, the partition contour area information is processed based on a preset processing rule to generate fluorescence signal information, including: obtaining a binary image of a preset contour, wherein the binary image of the preset contour is generated by creating a new canvas based on a fluorescent staining image and a partition contour image, and there is at least one binary image of the preset contour; processing the binary image of the fluorescence signal and the binary image of the preset contour to generate a coverage area of ​​the fluorescence signal in the binary image of the preset contour; processing the coverage area of ​​the fluorescence signal in the binary image of the preset contour to generate contour coordinate information; processing the contour coordinate information to generate minimum circle information of the coverage area of ​​the fluorescence signal in all binary images of the preset contours, wherein the minimum circle information includes radius information and center coordinate information of the minimum circle; processing the fluorescent pixel points to generate fluorescence signal distribution area information; processing the sum of the brightness of the fluorescent pixel points to generate fluorescence signal intensity; processing the minimum circle information and the fluorescence signal distribution area information to generate the discreteness of the fluorescence signal.

[0013] In one embodiment of the present application, the partition contour area information is processed based on a preset processing rule to generate fluorescent signal information, and further includes: the target fluorescent substance recognition model includes a calculation formula for obtaining the minimum circle radius and the center coordinates, and the calculation formula is:

[0014]

[0015] (X_Cernter, Y_Center)=Centroid_Coordinates;

[0016] Among them, Area min_circle is the area of ​​the smallest circle, Centroid_Coordinates are the centroid coordinates of the fluorescence signal;

[0017] The target fluorescent substance recognition model includes a calculation formula for obtaining the fluorescent signal distribution area, and the calculation formula is:

[0018] Area fluorescence =∑(Pixel_Value×Pixel_Area);

[0019] Among them, Pixel_Value is the fluorescence intensity of the pixel, and Pixel_Area is the area of ​​a single pixel;

[0020] The target fluorescent substance recognition model includes a calculation formula for obtaining the fluorescence signal intensity, and the calculation formula is:

[0021] Intensity fluorescence =∑Pixel_Value;

[0022] Among them, Pixel_Value is the fluorescence intensity of the pixel;

[0023] The target fluorescent substance recognition model includes a calculation formula for obtaining the discreteness of the fluorescent signal, and the calculation formula is:

[0024]

[0025] Among them, Min_Circle_Area is the area of ​​the smallest circle covering all fluorescent signals, and Fluorescence_Area is the total area of ​​fluorescent signals.

[0026] Another aspect of the present application is an immunofluorescence image analysis device, characterized in that it includes: an acquisition module, used to acquire target biological sample data and a training sample set matching the target biological sample data; acquire a preset fluorescent substance recognition model matching the fluorescent staining image; a processing module, used to process the target biological sample data to generate a fluorescent staining image, wherein the fluorescent staining image is an image of a single fluorescent channel; preprocess the training sample set to generate a training sample set with target feature data; based on the training sample set with target feature data, process the preset fluorescent substance recognition model to generate a target fluorescent substance recognition model; based on the target fluorescent substance recognition model, process the fluorescent staining image to generate original partition image information, wherein the original partition image information includes an original partition contour image and a fluorescent signal image; process the original partition image information to generate fluorescent signal information, wherein the fluorescent signal information includes a fluorescent signal distribution area, a fluorescent signal intensity, and a discreteness of the fluorescent signal; process the fluorescent signal distribution area, the fluorescent signal intensity, and the discreteness of the fluorescent signal to generate a fluorescent image spatial distribution feature.

[0027] According to another aspect of the present application, an electronic device is provided, characterized in that it comprises: a first processor; and a memory for storing executable instructions of the first processor; wherein the first processor is configured to execute the above-mentioned immunofluorescence image analysis method by executing the executable instructions.

[0028] According to another aspect of the present application, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a second processor, the above-mentioned immunofluorescence image analysis method is implemented.

[0029] According to another aspect of the present application, a computer program product is provided, including a computer program, characterized in that when the computer program is executed by a third processor, the above-mentioned immunofluorescence image analysis method is implemented.

[0030] The present application provides an immunofluorescence image analysis method and related equipment, in which a server performs statistical analysis of quantitative characteristic values ​​on a specific area circled in a single-channel fluorescence image, and provides various characteristic values ​​such as the area of ​​complex fluorescence signal distribution under different thresholds, the minimum circle area containing all fluorescence signals, the minimum circle radius and the coordinates of the center of the circle, and the discreteness of the fluorescence signal distribution (minimum circle area / fluorescence signal distribution area), thereby further providing quantitative clues for the description and classification of the complex fluorescence signal distribution.

[0031] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 A flow chart showing an immunofluorescence image analysis method provided by an embodiment of the present application;

[0033] Figure 2 A schematic diagram of the structure of an immunofluorescence image analysis device provided in one embodiment of the present application is shown;

[0034] Figure 3 A schematic diagram of the structure of an electronic device provided by an embodiment of the present application is shown;

[0035] Figure 4 A schematic diagram of a storage medium provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0036] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0037] Combine the following Figure 1 To describe the analysis method of immunofluorescence images according to an exemplary embodiment of the present application. It should be noted that the following application scenarios are only shown for the purpose of facilitating understanding of the spirit and principle of the present application, and the embodiments of the present application are not limited in this regard. On the contrary, the embodiments of the present application can be applied to any applicable scenario.

[0038] In one embodiment, the present application also proposes an immunofluorescence image analysis method and related equipment. Figure 1 The following schematically shows a flow chart of an immunofluorescence image analysis method according to an embodiment of the present application. Figure 1 As shown, the method is applied to a server, comprising:

[0039] S101, obtaining target biological sample data and a training sample set matching the target biological sample data.

[0040] In one embodiment, the target biological sample data is mainly for obtaining protein molecules or RNA molecules, which are obtained through cell culture, tissue sections and biological fluids.

[0041] Cell culture can be obtained in the following ways: select appropriate cell lines or primary cells according to the research purpose, prepare appropriate culture medium, including nutrients, growth factors and antibiotics. Inoculate cells into culture containers, such as culture dishes, culture bottles or multi-well plates, maintain appropriate temperature, pH value, CO2 concentration and humidity, and regularly pass cells according to cell growth to maintain cell activity and prevent aging. Regularly perform cell counting and viability tests to ensure healthy cell growth.

[0042] Tissue sections can be obtained in the following ways: collect tissue samples from an organism, such as through surgery or biopsy, and fix the tissue samples with an appropriate fixative (such as formaldehyde) to preserve the tissue structure. Dehydrate through a series of alcohol solutions with increasing concentrations, and then use a transparent agent (such as xylene) to embed the dehydrated tissue samples in an appropriate medium, such as paraffin or OCT (optimal cutting temperature) compound. Use a microtome to cut the embedded tissue into thin slices, usually between 4-10 microns in thickness, and stain the tissue sections, such as HE staining or immunohistochemical staining, to highlight specific cell structures or molecules.

[0043] For biofluid collection, it can be obtained in the following ways: select appropriate biofluids according to the research purpose, such as blood, urine, cerebrospinal fluid or cell culture supernatant, collect biofluids using sterile techniques, and pay attention to avoid contamination. The collected biofluids are divided into appropriate containers and stored at low temperatures as needed. The biofluids are centrifuged to separate cells, cell fragments and other components. Specific biomolecules, such as DNA, RNA or protein, are extracted as needed.

[0044] For quality control collection, it can be obtained by the following methods: quality control of extracted biomolecules, such as measuring concentration, purity and integrity. In the process of collecting and processing biological samples, laboratory safety regulations and bioethical principles need to be strictly followed to ensure the quality of samples and the reliability of research. The collection and processing methods of samples may be adjusted according to specific experimental design and analysis requirements.

[0045] Additionally, training data matching the target biological sample is collected, which may include images of known protein or RNA molecules and corresponding labels or annotations.

[0046] S102, processing the target biological sample data to generate a fluorescent staining image.

[0047] In one embodiment, after immunofluorescence staining of protein molecules or fluorescent in situ hybridization staining of RNA molecules, a single fluorescent channel image can be obtained by shooting under a fluorescent microscope. Among them, fluorescent staining is a commonly used biological sample labeling technology, which is used to observe and analyze specific molecules in cells or tissues under a microscope. The following are the steps for processing the target biological sample data and generating fluorescent staining images: First, fix the cell culture or tissue slices on a slide, treat the sample with a permeabilizing agent (such as TritonX-100) to increase the penetration of the antibody or probe, and use a protein blocker (such as bovine serum albumin or skim milk powder solution) to block nonspecific binding. The specific primary antibody (first antibody) is bound to the target protein molecule, the unbound primary antibody is washed to remove the secondary antibody, and a secondary antibody (second antibody) labeled with a fluorophore is added to specifically bind to the primary antibody.

[0048] Use fluorescently labeled nucleic acid probes to hybridize with target RNA molecules, wash to remove unbound probes, and after staining, wash the sample thoroughly to remove unbound antibodies or probes. Use anti-fluorescence fading agents (such as DAPI) to label cell nuclei, then use mounting media to seal the slides, use a fluorescence microscope to photograph the sample, and select appropriate fluorescence filters to capture the signal of specific fluorophores. Obtain images of a single fluorescent channel, each channel corresponding to a different fluorescent marker, and images of a combination of multiple fluorescent channels to simultaneously observe the distribution of multiple molecules. Fluorescent staining images are usually stored in BGR (blue-green-red) format. Convert BGR images to RGB (red-green-blue) format for viewing and analysis on standard display devices. Store RGB image files as backup to ensure data security and accessibility. Use image analysis software to perform quantitative analysis of fluorescent images, including fluorescence intensity measurement, co-localization analysis, and morphology measurement.

[0049] During fluorescent staining and image acquisition, care should be taken to avoid photobleaching and over-staining to ensure image quality and the accuracy of experimental results. In addition, appropriate software and algorithms should be used during image processing and analysis to improve the reliability and reproducibility of the results.

[0050] S103, obtaining a preset fluorescent substance recognition model that matches the fluorescent staining image.

[0051] In one embodiment, a suitable preset model is selected according to the characteristics of the target biological sample. This model can identify and analyze specific substances in fluorescent staining images, such as proteins or RNA molecules, and can solve different types of fluorescence microscopy image enhancement problems, including image super-resolution, isotropic reconstruction, 3D denoising, image projection, and process reconstruction. It shows excellent generalization and versatility in cross-modal datasets and cross-task experiments, significantly improves image enhancement accuracy, breaks through the limits of fluorescence microscopy imaging in multiple dimensions, and reveals clear cell structures of living samples. Fluorescent staining images can be effectively used to study the characteristics of biological samples, providing support for further biological research and potential clinical applications.

[0052] S104, preprocessing the training sample set to generate a training sample set with target feature data.

[0053] In one embodiment, feature extraction is performed on the training sample set to determine the original feature library, and each feature data set is divided according to the original feature library to generate a training data set and a verification data set. The classifier is used to predict each verification data set divided from the original feature library to determine the prediction result. A preset algorithm is used to divide each training data set into the original feature library for training to obtain a verification set class prediction result, and target feature data is generated based on the prediction result and the verification set class prediction result, and the target feature data is used to characterize the spatial distribution characteristics of the fluorescence image.

[0054] Specifically, it is necessary to extract features from the training sample set. This step involves identifying basic attributes in the image, such as fluorescence intensity, morphological features, texture features, etc. These features together constitute the original feature library, which lays the foundation for subsequent model training. According to the original feature library, the dataset is further divided into a training dataset and a validation dataset. The training dataset is used for preliminary training of the model, while the validation dataset is used to evaluate model performance and perform hyperparameter tuning. The validation dataset is predicted using a classifier to determine the prediction results of the model. This step helps to evaluate the accuracy and generalization ability of the model. The preset algorithm is used to train on the training dataset to obtain the class prediction results of the validation set. This process may involve multiple machine learning or deep learning algorithms to identify the model that can most accurately predict the target features.

[0055] According to the prediction results and the prediction results of the validation set, target feature data are generated. These data are used to characterize the spatial distribution characteristics of the fluorescence image, such as the distribution area, intensity, and discreteness of the fluorescence signal. Through continuous iterative optimization, a model that can accurately identify and quantify the fluorescence signal is finally determined. This model can be applied to new fluorescence staining images to provide localization and quantitative information about the target protein molecule or RNA molecule. The fluorescence staining images are analyzed, focusing on the distribution characteristics of the fluorescence signal, such as aggregated distribution or diffuse distribution. These characteristics are crucial to understanding the function of the target molecule. Through the above steps, researchers can have a deeper understanding of the characteristics of biological samples and provide support for biological research and potential clinical applications.

[0056] S105 , processing a preset fluorescent substance recognition model based on the training sample set with target feature data to generate a target fluorescent substance recognition model.

[0057] In one embodiment, a training sample set with target feature data is divided into a training set for training a preset fluorescent substance recognition model and a verification set for verifying the preset fluorescent substance recognition model based on a preset ratio, and multiple data groups are extracted from the training set, wherein each data group contains a preset number of data samples, wherein at least one data sample includes target feature data. The preset fluorescent substance recognition model is trained based on the data samples in the multiple data groups to generate a trained fluorescent substance recognition model, and the trained fluorescent substance recognition model is processed based on the verification set to generate a verification result, and if the data samples containing the target feature data in the verification result are fluorescent pixels and the brightness of the fluorescent pixels, the trained fluorescent substance recognition model is used as the target fluorescent substance recognition model.

[0058] Specifically, the training sample set with target feature data is divided into a training set and a validation set according to a preset ratio (for example, the common 70% training set and 30% validation set), and multiple data sets are extracted from the training set to ensure that each data set contains a predetermined number of samples and at least one sample with target feature data. These data sets are used to train the preset fluorescent substance recognition model. This process may include multiple iterations to optimize the model parameters, evaluate the trained model using the validation set, and generate validation results. This step is the key to verifying the generalization ability of the model.

[0059] Analyze the validation results. If the model can effectively identify fluorescent pixels and their brightness, and the results contain target feature data, the model has good predictive performance. If the validation results are satisfactory, the trained model is used as the target fluorescent substance recognition model for subsequent fluorescence image analysis. Depending on the validation results, it may be necessary to return and further optimize the model, such as adjusting the model structure, training more iterations, or using different features. Apply the final confirmed model to new fluorescent staining images to identify and quantify fluorescent signals. This process is iterative and may require multiple adjustments and optimizations to achieve the best model performance. In addition, some techniques such as cross-validation and regularization may be used in the model training and validation process to improve the stability and generalization ability of the model.

[0060] In another embodiment, the training set is divided based on the target feature data to generate a number of number class samples, obtain any number of number class samples in the training set that are less than a preset threshold, and generate adjacent samples based on the distance between any number of number class samples that are less than the preset threshold and other number class samples of the same category that are less than the preset threshold, wherein the adjacent samples include a preset number of number class samples that are less than the preset threshold. Based on the number of samples of each number class in the training set, a sampling ratio is determined, based on the sampling ratio, a sampling ratio is determined, and based on the sampling ratio, adjacent samples are sampled to generate a preset number of sampled samples, and multiple data groups are generated based on any class of samples and each sampled sample.

[0061] Specifically, according to the target feature data, the training set is divided into a number of class samples, that is, the data is divided into different categories or groups based on the features, and the class samples whose number of samples is less than the preset threshold are found from each category. These samples may represent minority classes and require special attention to avoid model bias. For class samples whose number is less than the threshold, adjacent samples are found based on the distance between them (for example, Euclidean distance or Manhattan distance). This usually involves selecting other minority samples that are closest to these minority samples to enhance the model's recognition ability for minority classes.

[0062] According to the number of samples of each category in the training set, a sampling ratio is determined. This ratio is used to adjust the representativeness of samples of each category in the training set. According to the sampling ratio, the sampling ratio is further determined. The sampling ratio guides how to extract samples from each category to ensure a more balanced distribution of categories in the training set. According to the sampling ratio, adjacent samples are sampled to generate a preset number of sampled samples. The purpose of this step is to create a more balanced training set and improve the model's ability to recognize minority categories.

[0063] Based on any class of samples and each sampled sample, multiple data sets are generated. These data sets will be used for model training to improve the generalization and robustness of the model, aiming to improve the model's recognition ability for minority categories through intelligent sampling and data enhancement techniques, thereby achieving better performance when processing unbalanced data sets, which is particularly important for fields such as medical image analysis, as some pathological features may appear less frequently in images.

[0064] S106: Process the fluorescent staining image based on the target fluorescent substance recognition model to generate original partition image information.

[0065] In one embodiment, the original partition image information includes an original partition outline image and a fluorescent signal image. The fluorescent staining image is processed based on the target fluorescent substance recognition model, and the original partition image information is generated and a quantitative eigenvalue statistical analysis is performed, as follows: the single-channel fluorescent staining image is input into the target fluorescent substance recognition model, and the input fluorescent image is processed using the target fluorescent substance recognition model to identify and quantify the fluorescent signal, and the processing result includes an original partition outline image and a fluorescent signal image. The original partition outline image shows the boundary of the circled area, while the fluorescent signal image shows the distribution of the fluorescent substance in a specific area.

[0066] Analyze the fluorescence signal image in a specific area and extract quantitative feature values. These features include but are not limited to: the total area of ​​the fluorescence signal, the distribution of the fluorescence intensity, including the average intensity, the coefficient of variation of the intensity, etc., and the morphological characteristics of the fluorescence signal, such as shape, size, irregularity, etc.

[0067] Image analysis algorithms, such as threshold segmentation, edge detection, and morphological operations, are used to extract contour features from the original partitioned contour image, and the extracted features are quantified, such as calculating the area, perimeter, and shape descriptors (such as circularity and ellipticity) of the fluorescent area. All extracted quantitative features are integrated into a data structure, such as a feature vector or a data frame, for further analysis. The biological significance of the fluorescent substance is interpreted based on the extracted feature values, which may involve associations with known biological processes or disease states. In order to better understand the distribution and characteristics of the fluorescent signal, the results are visualized, such as generating a heat map of the fluorescence intensity, a distribution map of the fluorescence signal, etc. This process is highly automated and can greatly improve the efficiency and accuracy of fluorescence microscopy image analysis, which is of great value for biomedical research.

[0068] S107, processing the original partition image information to generate fluorescence signal information.

[0069] In one embodiment, the original partition image information is preprocessed to generate partition image information in a target format, wherein the partition image information in the target format includes a partition outline image and a binary image of a fluorescence signal, and the binary image of a fluorescence signal includes fluorescent pixels and the sum of the brightness of the fluorescent pixels. Specifically, the original partition image information is preprocessed to generate image data in a target format, including: a partition outline image: showing the boundary of the circled area. A binary image of a fluorescence signal: through threshold processing, the pixels of the fluorescence signal are divided into two categories (for example, fluorescence and non-fluorescence). In the binary image, the fluorescent pixels are represented as white (or pixels with higher brightness), and the non-fluorescent pixels are represented as black (or pixels with lower brightness). The sum of the brightness of the fluorescent pixels is calculated, which can be used as a measure of the fluorescence intensity.

[0070] The partition contour image is grayed to generate a grayscale image, the grayscale image is binarized to generate the coordinates of the outermost contour endpoints, the partition contour area information is generated based on the coordinates of the outermost contour endpoints, the partition contour area information is processed based on preset processing rules to generate fluorescence signal information, and the fluorescence signal information includes the fluorescence signal distribution area, the fluorescence signal intensity and the discreteness of the fluorescence signal.

[0071] Specifically, the partition outline image is converted into a grayscale image for subsequent processing. Grayscaling is the process of converting a color or multi-channel image into a single-channel image. The grayscale image is binarized to highlight the outermost contour. This step usually involves selecting a threshold to divide the pixels in the image into foreground and background, and extracting the endpoint coordinates of the outermost contour from the binarized image, which define the boundaries of the region of interest. Using the extracted contour endpoint coordinates, the area information of the partition outline is calculated, which can provide data on the size of the fluorescent signal distribution area. The partition outline area information is further processed according to preset processing rules (e.g., morphological operations, signal filtering, etc.). After processing through the above steps, fluorescent signal information is generated, which may include fluorescence intensity, distribution uniformity, morphological characteristics, etc. The processed fluorescent signal is quantified, and key quantitative features such as the mean, median, and variance of the fluorescence intensity are extracted. All extracted features are recorded and statistically analyzed for further biological interpretation and research. Through this series of steps, researchers can extract valuable quantitative information from fluorescent staining images and then analyze and study the specific characteristics of biological samples. This information is of great significance for understanding cellular processes, disease mechanisms, and drug effects.

[0072] In another embodiment, first, extract the binary mask of the fluorescence signal on the image, set the upper and lower thresholds according to the selected fluorescence signal range, clear the exposed white area, extract the pixels within the threshold range on the image and set them to white, and set the remaining pixel values ​​to black. Secondly, smooth the extracted binary mask result of the fluorescence signal, define the size of the kernel, iterate two rounds of expansion kernel corrosion to ensure smoothness. Then, extract the color according to the manually circled area border and convert it into a grayscale image for use, binarize the grayscale image, find the contour, and obtain the outermost contour and store the coordinates of the contour endpoints. Finally, calculate the contour area of ​​each area, filter out the contours with an area less than 1000, count the remaining number of contours as n, sort all contour areas, and extract the index values ​​of the n contours with the largest area.

[0073] In another embodiment, the partition contour area information is processed based on a preset processing rule to generate fluorescence signal information, specifically including: obtaining a binary image of a preset contour, wherein the binary image of the preset contour is generated by creating a new canvas based on a fluorescent staining image and a partition contour image, and there is at least one binary image of the preset contour, processing the binary image of the fluorescence signal and the binary image of the preset contour to generate the coverage area of ​​the fluorescence signal in the binary image of the preset contour. Processing the coverage area of ​​the fluorescence signal in the binary image of the preset contour to generate contour coordinate information, processing the contour coordinate information to generate the minimum circle information of the coverage area of ​​the fluorescence signal in all binary images of the preset contour, wherein the minimum circle information includes the radius information and the center coordinate information of the minimum circle. Processing the fluorescence pixel points to generate the fluorescence signal distribution area information, processing the sum of the brightness of the fluorescence pixel points to generate the fluorescence signal intensity, and processing the minimum circle information and the fluorescence signal distribution area information to generate the discreteness of the fluorescence signal.

[0074] Specifically, a blank canvas of the same size as the image can be created, and a single contour can be drawn on the canvas to generate a binary image, and the interior of the contour is completely filled. Secondly, a bitwise AND operation is performed on the binary image of the fluorescence signal and the binary image of the contour to generate a binary image in which the area covered by the fluorescence signal in the contour is white and the rest is black. Then, the contour of the area covered by the fluorescence signal in the above contour is found to obtain the contour coordinates, and the minimum circle covering all the fluorescent areas in the contour is calculated according to the coordinates to obtain the radius and center coordinates of the minimum circle. And the minimum circle is drawn on the binary image obtained above according to the center and radius. Finally, a black image with the same height, width and format as the image is created, and the minimum circle filled with white (255,255,255) is drawn on it according to the obtained minimum circle radius and center coordinates. The bitwise AND operation is performed on the pixel area with the same non-zero pixel position as the original image, and the specific fluorescence channel value after the operation is extracted. After flattening into a one-dimensional array, the pixel values ​​below the threshold are filtered out according to the selected threshold and the number of occurrences of each pixel value is calculated as the weight. The overall fluorescence intensity within the contour is obtained by weighted statistics based on the pixel value and the number of occurrences, and various characteristic values ​​such as the area of ​​the smallest circle and the discreteness of the fluorescence signal distribution are calculated.

[0075] In addition, for the analysis of a single image, create a file upload control where users can upload files. Create a slider to adjust the fluorescence threshold selection and define the minimum, maximum, and default values ​​of the slider. Users can drag the slider to adjust the threshold. Then, after the user completes the operation, the application will execute the above modules and display the boundary range map of the manually selected specific area, the fluorescence signal coverage map in the specific area, the original image, and the image superimposed with the minimum circle and signal range, and display the results of various feature values ​​on the application interface.

[0076] In another embodiment, the target fluorescent substance recognition model includes a calculation formula for obtaining the minimum circle radius and the coordinates of the circle center, and the calculation formula is:

[0077]

[0078] (X_Cernter, Y_Center)=Centroid_Coordinates;

[0079] Among them, Area min_circle is the area of ​​the smallest circle, Centroid_Coordinates are the centroid coordinates of the fluorescence signal;

[0080] The target fluorescent substance recognition model includes a calculation formula for obtaining the distribution area of ​​the fluorescent signal, which is:

[0081] Areafluorescence =∑(Pixel_Value×Pixel_Area);

[0082] Among them, Pixel_Value is the fluorescence intensity of the pixel, and Pixel_Area is the area of ​​a single pixel;

[0083] The target fluorescent substance recognition model includes a calculation formula for obtaining the fluorescence signal intensity, which is:

[0084] Intensity fluorescence =∑Pixel_Value;

[0085] Among them, Pixel_Value is the fluorescence intensity of the pixel;

[0086] The target fluorescent substance recognition model includes a calculation formula for obtaining the discreteness of the fluorescent signal, which is:

[0087]

[0088] Among them, Min_Circle_Area is the area of ​​the smallest circle covering all fluorescent signals, and Fluorescence_Area is the total area of ​​fluorescent signals.

[0089] S108, processing the fluorescence signal distribution area, the fluorescence signal intensity and the discreteness of the fluorescence signal to generate a fluorescence image spatial distribution feature.

[0090] In one embodiment, the spatial distribution feature processing of the fluorescence image generally involves the following key steps: preprocessing the original fluorescence image to improve the image quality and enhance the signal. This may include operations such as denoising, background correction, and signal enhancement. For example, a self-supervised approach can be used to remove noise from the fluorescence image, extract training pairs through a spatial redundant sampling strategy, and use a lightweight spatiotemporal Transformer architecture to capture long-range dependencies and high-resolution features to restore high-frequency information. Extract key features of the fluorescence signal, such as fluorescence intensity, distribution area, etc. from the preprocessed image. This can be achieved through methods such as image segmentation and threshold processing to distinguish between fluorescence signals and background. For example, when using ImageJ software for average fluorescence intensity detection, the appropriate area can be selected by adjusting the threshold, and the fluorescence intensity of a specific area can be calculated.

[0091] The extracted features are quantitatively analyzed to calculate the statistical parameters of the fluorescence signal, such as the mean fluorescence intensity, peak intensity, uniformity of distribution, etc. These parameters can provide a quantitative description of the spatial distribution of the fluorescent substance. All the extracted quantitative features are integrated and further analyzed, such as principal component analysis (PCA) or cluster analysis, to identify different fluorescence patterns or groups. The discreteness of the fluorescence signal is calculated, which may involve the analysis of the variability of the fluorescence intensity distribution. For example, in flow cytometry measurements, the variability of the fluorescence intensity distribution may be affected by cell size and cell cycle progression, which needs to be considered in the analysis. Based on the results of quantitative analysis and discreteness calculation, the spatial distribution characteristics of the fluorescence image are interpreted to reveal the biological significance or function of the fluorescent substance.

[0092] The present application comprises the following steps: obtaining target biological sample data and a training sample set matching the target biological sample data by a server; processing the target biological sample data to generate a fluorescent staining image, wherein the fluorescent staining image is an image of a single fluorescent channel; obtaining a preset fluorescent substance recognition model matching the fluorescent staining image; performing feature extraction on the training sample set to determine an original feature library; dividing each feature data set according to the original feature library to generate a training data set and a verification data set; using a classifier to predict each verification data set divided from the original feature library to determine a prediction result; using a preset algorithm to train each training data set divided from the original feature library to obtain a verification set class prediction result; generating target feature data based on the prediction result and the verification set class prediction result, wherein the target feature data is used to characterize the spatial distribution characteristics of the fluorescent image; dividing the training sample set with the target feature data into a training set for training a preset fluorescent substance recognition model and a verification set for verifying the preset fluorescent substance recognition model based on a preset ratio; dividing the training set based on the target feature data to generate a number of sample classes.

[0093] Obtain any number of samples of a class in the training set that are less than a preset threshold; generate adjacent samples based on the distance between any number of samples of a class that are less than a preset threshold and other number of samples of the same class that are less than a preset threshold, wherein the adjacent samples include a preset number of samples of any class that are less than a preset threshold. Determine the sampling ratio based on the number of samples of each class in the training set; determine the sampling ratio based on the sampling ratio; sample adjacent samples based on the sampling ratio to generate a preset number of sampled samples; generate multiple data groups based on any class of samples and each sampled sample, wherein each data group includes a preset number of data samples, wherein at least one data sample includes target feature data; train a preset fluorescent substance recognition model based on the data samples in the multiple data groups to generate a trained fluorescent substance recognition model; process the trained fluorescent substance recognition model based on the validation set to generate a validation result; if the data samples containing the target feature data in the validation result are fluorescent pixels and the brightness of the fluorescent pixels, the trained fluorescent substance recognition model is used as the target fluorescent substance recognition model. The fluorescent staining image is processed based on the target fluorescent substance recognition model to generate original partition image information, wherein the original partition image information includes the original partition outline image and the fluorescent signal image, and the original partition image information is preprocessed to generate the partition image information in the target format, wherein the partition image information in the target format includes the partition outline image and the binary image of the fluorescent signal, and the binary image of the fluorescent signal includes the fluorescent pixel points and the sum of the brightness of the fluorescent pixel points. The partition outline image is grayed to generate a gray image, and the gray image is binarized to generate the outermost contour endpoint coordinates, and the partition outline area information is generated based on the outermost contour endpoint coordinates, and the binary image of the preset contour is obtained, wherein the binary image of the preset contour is generated by creating a new canvas based on the fluorescent staining image and the partition outline image, and there is at least one binary image of the preset contour.

[0094] The binary image of the fluorescence signal and the binary image of the preset contour are processed to generate the coverage area of ​​the fluorescence signal in the binary image of the preset contour, the coverage area of ​​the fluorescence signal in the binary image of the preset contour is processed to generate contour coordinate information, the contour coordinate information is processed to generate the minimum circle information of the coverage area of ​​the fluorescence signal in all binary images of the preset contour, wherein the minimum circle information includes the radius information and the center coordinate information of the minimum circle, the fluorescence pixel points are processed to generate the fluorescence signal distribution area information, the sum of the brightness of the fluorescence pixel points is processed to generate the fluorescence signal intensity, the minimum circle information and the fluorescence signal distribution area information are processed to generate the discreteness of the fluorescence signal, wherein the fluorescence signal information includes the fluorescence signal distribution area, the fluorescence signal intensity and the discreteness of the fluorescence signal. The fluorescence signal distribution area, the fluorescence signal intensity and the discreteness of the fluorescence signal are processed to generate the spatial distribution characteristics of the fluorescence image. Statistical analysis of quantitative characteristic values ​​is performed on specific areas circled in single-channel fluorescence images, providing various characteristic values ​​such as the area of ​​complex fluorescence signal distribution under different thresholds, the minimum circle area containing all fluorescence signals, the minimum circle radius and circle center coordinates, and the discreteness of fluorescence signal distribution (minimum circle area / fluorescence signal distribution area), further providing quantitative clues for the description and classification of complex fluorescence signal distribution.

[0095] In one embodiment, if Figure 2 As shown, the present application also provides an immunofluorescence image analysis device, comprising:

[0096] The acquisition module 201 is used to acquire target biological sample data and a training sample set matching the target biological sample data; and acquire a preset fluorescent substance recognition model matching the fluorescent staining image;

[0097] The processing module 202 is used to process the target biological sample data to generate a fluorescent staining image, wherein the fluorescent staining image is an image of a single fluorescent channel; pre-process the training sample set to generate a training sample set with target feature data; process a preset fluorescent substance recognition model based on the training sample set with target feature data to generate a target fluorescent substance recognition model; process the fluorescent staining image based on the target fluorescent substance recognition model to generate original partition image information, wherein the original partition image information includes an original partition contour image and a fluorescent signal image; process the original partition image information to generate fluorescent signal information, wherein the fluorescent signal information includes a fluorescent signal distribution area, a fluorescent signal intensity, and a discreteness of the fluorescent signal; process the fluorescent signal distribution area, the fluorescent signal intensity, and the discreteness of the fluorescent signal to generate a fluorescent image spatial distribution feature.

[0098] In another embodiment of the present application, the processing module 202 is configured to pre-process the training sample set to generate a training sample set with target feature data, including:

[0099] Extracting features from the training sample set to determine an original feature library;

[0100] Divide each feature data set according to the original feature library to generate a training data set and a verification data set;

[0101] Use the classifier to divide the original feature library into various verification data sets for prediction and determine the prediction results;

[0102] Use the preset algorithm to divide each training data set in the original feature library for training, and obtain the prediction results of the validation set class;

[0103] According to the prediction results and the prediction results of the validation set, target feature data is generated, and the target feature data is used to characterize the spatial distribution characteristics of the fluorescence image.

[0104] In another embodiment of the present application, the processing module 202 is configured to process the preset fluorescent substance recognition model based on the training sample set with target feature data to generate a target fluorescent substance recognition model, including:

[0105] Dividing the training sample set with target feature data into a training set for training the preset fluorescent substance recognition model and a verification set for verifying the preset fluorescent substance recognition model based on a preset ratio;

[0106] Extracting a plurality of data groups from the training set, wherein each data group comprises a preset number of data samples, wherein at least one data sample comprises target feature data;

[0107] Training the preset fluorescent substance recognition model based on the data samples in the plurality of data groups to generate a trained fluorescent substance recognition model;

[0108] Processing the trained fluorescent substance recognition model based on the verification set to generate a verification result;

[0109] If the data samples containing the target characteristic data in the verification result are fluorescent pixels and the brightness of the fluorescent pixels, the trained fluorescent substance recognition model is used as the target fluorescent substance recognition model.

[0110] In another embodiment of the present application, the processing module 202 is configured to extract multiple data sets from the training set, including:

[0111] Dividing the training set based on target feature data to generate a number of numerical samples;

[0112] Obtain any number of samples in the training set whose number is less than a preset threshold;

[0113] Generate adjacent samples based on the distance between any number of samples of the number class less than the preset threshold and other number class samples of the same category less than the preset threshold, wherein the adjacent samples include a preset number of samples of any number class less than the preset threshold;

[0114] Determining a sampling ratio based on the number of samples of each category in the training set;

[0115] Based on the sampling ratio, determining a sampling ratio;

[0116] Sampling the adjacent samples based on the sampling ratio to generate a preset number of sampling samples;

[0117] Based on any class of samples and each sampling sample, multiple groups of data are generated.

[0118] In another embodiment of the present application, the processing module 202 is configured to process the original partition image information to generate fluorescence signal information, including:

[0119] Preprocessing the original subarea image information to generate subarea image information in a target format, wherein the subarea image information in the target format includes a subarea contour image and a binary image of a fluorescent signal, and the binary image of a fluorescent signal includes fluorescent pixels and the sum of the brightness of the fluorescent pixels;

[0120] grayscale the partition outline image to generate a grayscale image;

[0121] Binarizing the grayscale image to generate the coordinates of the outermost contour endpoints;

[0122] Generate partition contour area information based on the outermost contour endpoint coordinates;

[0123] The partition contour area information is processed based on preset processing rules to generate fluorescent signal information.

[0124] In another embodiment of the present application, the processing module 202 is configured to process the partition contour area information based on a preset processing rule to generate fluorescent signal information, including:

[0125] Acquire a binary image of a preset contour, wherein the binary image of the preset contour is generated by the newly created canvas based on the fluorescent staining image and the partition contour image, and there is at least one binary image of the preset contour;

[0126] Processing the binary image of the fluorescence signal and the binary image of the preset contour to generate a coverage area of ​​the fluorescence signal in the binary image of the preset contour;

[0127] Processing the coverage area of ​​the fluorescence signal in the binary image of the preset contour to generate contour coordinate information;

[0128] Processing the contour coordinate information to generate minimum circle information of the coverage area of ​​the fluorescence signal in the binary image of all preset contours, wherein the minimum circle information includes radius information and center coordinate information of the minimum circle;

[0129] Processing the fluorescent pixel points to generate fluorescent signal distribution area information;

[0130] Processing the sum of the brightness of the fluorescent pixels to generate a fluorescent signal intensity;

[0131] The minimum circle information and the fluorescence signal distribution area information are processed to generate the discreteness of the fluorescence signal.

[0132] The present application embodiment provides an electronic device, such as Figure 3 As shown, the electronic device 3 includes a first processor 300, a memory 301, a bus 302 and a communication interface 303. The first processor 300, the communication interface 303 and the memory 301 are connected via the bus 302; the memory 301 stores a computer program that can be run on the first processor 300, and when the first processor 300 runs the computer program, the immunofluorescence image analysis method provided in any of the aforementioned embodiments of the present application is executed.

[0133] The memory 301 may include a high-speed random access memory (RAM), and may also include a non-volatile memory, such as at least one disk memory. The communication connection between the system network element and at least one other network element is realized through at least one communication interface 303 (which may be wired or wireless), and the Internet, wide area network, local area network, metropolitan area network, etc. may be used.

[0134] The bus 302 may be an ISA bus, a PCI bus, or an EISA bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. The memory 301 is used to store a program, and the first processor 300 executes the program after receiving an execution instruction. The immunofluorescence image analysis method disclosed in any of the embodiments of the present application may be applied to the first processor 300, or implemented by the first processor 300.

[0135] The first processor 300 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit or software instructions in the first processor 300. The above-mentioned first processor 300 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a readily available programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiments of the present application can be embodied as a hardware decoding processor to be executed, or a combination of hardware and software modules in the decoding processor to be executed. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 301, and the first processor 300 reads the information in the memory 301 and completes the steps of the above method in combination with its hardware.

[0136] The electronic device provided in the above-mentioned embodiment of the present application and the immunofluorescence image analysis method provided in the embodiment of the present application are based on the same inventive concept and have the same beneficial effects as the method adopted, run or implemented by the application program stored therein.

[0137] The present application embodiment provides a computer-readable storage medium, such as Figure 4 As shown, the computer-readable storage medium stores 401 a computer program, and when the computer program is read and executed by the second processor 402, the aforementioned immunofluorescence image analysis method is implemented.

[0138] The technical solution of the embodiment of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for enabling an electronic device (which may be an air conditioner, a refrigeration device, a personal computer, a server, or a network device, etc.) or a processor to execute all or part of the steps of the method described in the embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a ROM, a RAM, a magnetic disk, or an optical disk.

[0139] The computer-readable storage medium provided in the above-mentioned embodiments of the present application and the immunofluorescence image analysis method provided in the embodiments of the present application are based on the same inventive concept and have the same beneficial effects as the method adopted, run or implemented by the application program stored therein.

[0140] An embodiment of the present application provides a computer program product, including a computer program, wherein the computer program is executed by a third processor to implement the method described above.

[0141] The computer program product provided in the above-mentioned embodiments of the present application and the immunofluorescence image analysis method provided in the embodiments of the present application are based on the same inventive concept and have the same beneficial effects as the method adopted, run or implemented by the application program stored therein.

[0142] It should be noted that, in this application, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or still includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise one..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0143] Each embodiment in the present application is described in a related manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the analysis method, electronic device, electronic device, and readable storage medium embodiment for evaluating immunofluorescence images, since they are basically similar to the above-mentioned immunofluorescence image analysis method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the above-mentioned immunofluorescence image analysis method embodiment.

[0144] Although the present application is disclosed as above, the present application is not limited thereto. Any person skilled in the art can make various changes and modifications without departing from the spirit and scope of the present application. Therefore, the protection scope of the present application shall be subject to the scope defined by the claims.

Claims

1. A method for analyzing immunofluorescence images, characterized in that: include: Obtaining target biological sample data and a training sample set matching the target biological sample data; Processing the target biological sample data to generate a fluorescent staining image, wherein the fluorescent staining image is an image of a single fluorescent channel; Acquiring a preset fluorescent substance recognition model that matches the fluorescent staining image; Preprocessing the training sample set to generate a training sample set with target feature data; Processing a preset fluorescent substance recognition model based on the training sample set with target feature data to generate a target fluorescent substance recognition model; Processing the fluorescent staining image based on the target fluorescent substance recognition model to generate original partition image information, wherein the original partition image information includes an original partition contour image and a fluorescent signal image; Processing the original partition image information to generate fluorescence signal information includes: obtaining a binary image of a preset contour, wherein the binary image of the preset contour is generated by creating a new canvas based on a fluorescent staining image and a partition contour image, and there is at least one binary image of the preset contour; processing the binary image of the fluorescence signal and the binary image of the preset contour to generate a coverage area of ​​the fluorescence signal in the binary image of the preset contour, wherein the binary image of the fluorescence signal includes fluorescence pixels and the sum of the brightness of the fluorescence pixels; processing the coverage area of ​​the fluorescence signal in the binary image of the preset contour to generate contour coordinate information; processing the contour coordinate information to generate minimum circle information of the coverage area of ​​the fluorescence signal in all binary images of the preset contour, wherein the minimum circle information includes the radius information and the center coordinate information of the minimum circle; processing the fluorescence pixels to generate fluorescence signal distribution area information; processing the sum of the brightness of the fluorescence pixels to generate fluorescence signal intensity; processing the minimum circle information and the fluorescence signal distribution area information to generate the discreteness of the fluorescence signal; The fluorescence signal distribution area, fluorescence signal intensity and the discreteness of the fluorescence signal are processed to generate the spatial distribution characteristics of the fluorescence image.

2. The method according to claim 1, characterized in that The preprocessing of the training sample set to generate a training sample set with target feature data includes: Extracting features from the training sample set to determine an original feature library; Divide each feature data set according to the original feature library to generate a training data set and a verification data set; Use the classifier to divide the original feature library into various verification data sets for prediction and determine the prediction results; Use the preset algorithm to divide each training data set in the original feature library for training, and obtain the prediction results of the validation set class; According to the prediction results and the prediction results of the validation set, target feature data is generated, and the target feature data is used to characterize the spatial distribution characteristics of the fluorescence image.

3. The method according to claim 2, characterized in that The preset fluorescent substance recognition model is processed based on the training sample set with target feature data to generate a target fluorescent substance recognition model, including: Dividing the training sample set with target feature data into a training set for training the preset fluorescent substance recognition model and a verification set for verifying the preset fluorescent substance recognition model based on a preset ratio; Extracting a plurality of data groups from the training set, wherein each data group comprises a preset number of data samples, wherein at least one data sample comprises target feature data; Training the preset fluorescent substance recognition model based on the data samples in the plurality of data groups to generate a trained fluorescent substance recognition model; Processing the trained fluorescent substance recognition model based on the verification set to generate a verification result; If the data samples containing the target characteristic data in the verification result are fluorescent pixels and the brightness of the fluorescent pixels, the trained fluorescent substance recognition model is used as the target fluorescent substance recognition model.

4. The method according to claim 3, characterized in that Extracting multiple data sets from the training set includes: Dividing the training set based on target feature data to generate a number of numerical samples; Obtain any number of samples in the training set whose number is less than a preset threshold; Generate adjacent samples based on the distance between any number of samples of the number class less than the preset threshold and other number class samples of the same category less than the preset threshold, wherein the adjacent samples include a preset number of samples of any number class less than the preset threshold; Determining a sampling ratio based on the number of samples of each category in the training set; Based on the sampling ratio, determining a sampling ratio; Sampling the adjacent samples based on the sampling ratio to generate a preset number of sampling samples; Based on any class of samples and each sampling sample, multiple groups of data are generated.

5. The method according to claim 1, characterized in that Processing the original partition image information to generate fluorescence signal information also includes: Preprocessing the original subarea image information to generate subarea image information in a target format, wherein the subarea image information in the target format includes a subarea contour image and a binary image of a fluorescence signal; grayscale the partition outline image to generate a grayscale image; Binarizing the grayscale image to generate the coordinates of the outermost contour endpoints; The partition contour area information is generated based on the coordinates of the outermost contour endpoints.

6. The method according to claim 1, characterized in that Processing the partition contour area information based on a preset processing rule to generate fluorescence signal information also includes: The target fluorescent substance recognition model includes a calculation formula for obtaining the minimum circle radius and the coordinates of the circle center, and the calculation formula is: ； ； in, is the area of ​​the smallest circle, is the centroid coordinate of the fluorescence signal; The target fluorescent substance recognition model includes a calculation formula for obtaining the fluorescent signal distribution area, and the calculation formula is: ； in, is the fluorescence intensity of the pixel, is the area of ​​a single pixel; The target fluorescent substance recognition model includes a calculation formula for obtaining the fluorescence signal intensity, and the calculation formula is: ； in, is the fluorescence intensity of the pixel; The target fluorescent substance recognition model includes a calculation formula for obtaining the discreteness of the fluorescent signal, and the calculation formula is: ； in, is the area of ​​the smallest circle covering all fluorescent signals, is the total area of ​​the fluorescent signal.

7. An immunofluorescence image analysis device, characterized in that: For implementing the method according to claim 1, the device comprises: An acquisition module is used to acquire target biological sample data and a training sample set matching the target biological sample data; and to acquire a preset fluorescent substance recognition model matching the fluorescent staining image; A processing module is used to process the target biological sample data to generate a fluorescent staining image, wherein the fluorescent staining image is an image of a single fluorescent channel; pre-process the training sample set to generate a training sample set with target feature data; process a preset fluorescent substance recognition model based on the training sample set with target feature data to generate a target fluorescent substance recognition model; process the fluorescent staining image based on the target fluorescent substance recognition model to generate original partition image information, wherein the original partition image information includes an original partition contour image and a fluorescent signal image; process the original partition image information to generate fluorescent signal information, wherein the fluorescent signal information includes a fluorescent signal distribution area, a fluorescent signal intensity and a discreteness of the fluorescent signal; process the fluorescent signal distribution area, the fluorescent signal intensity and the discreteness of the fluorescent signal to generate a fluorescent image spatial distribution feature.

8. An electronic device, characterized in that: include: a first processor; and a memory for storing executable instructions of the first processor; Wherein, the first processor is configured to execute the immunofluorescence image analysis method according to any one of claims 1 to 6 by executing the executable instructions.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the second processor, the method for analyzing immunofluorescence images according to any one of claims 1 to 6 is implemented.

Images