A method and system for data acquisition and analysis of immunohistochemical staining machine

Abstract

This application relates to the field of data processing technology and discloses a method and system for data acquisition and analysis of an immunohistochemical staining machine. The method includes: acquiring multi-cell collaborative data through neighborhood cell staining intensity correlation analysis; constructing a signal transduction intensity matrix based on spatial distribution patterns; calculating the microenvironment influence coefficient using a cascade reaction prediction network; performing context correction on single-cell staining based on microenvironment influence; and finally performing microenvironment-aware analysis on the cell population to obtain staining analysis results based on cell interactions. This application solves the problem of lacking microenvironment context information in existing immunohistochemical analyses by constructing a computational model based on intercellular signal transduction cascade reactions, thereby improving the biological accuracy of staining assessment.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a method and system for data acquisition and analysis of an immunohistochemical staining machine. Background Technology

[0002] Existing immunohistochemical staining data acquisition and analysis methods are primarily based on a single-cell independent analysis framework. They acquire cell staining intensity data through optical detection systems and then use a fixed threshold classification method to categorize cells into different grades such as negative, weakly positive, and strongly positive. This technology is widely used in tumor pathology diagnosis for biomarker detection, such as the expression analysis of proteins like PD-L1, HER2, and Ki67, providing important information for clinical treatment selection. Traditional data processing workflows include image acquisition, cell segmentation, feature extraction, and classification assessment, with each step relatively independent and primarily focusing on the staining characteristics of individual cells.

[0003] Existing technologies suffer from significant technical limitations, primarily due to their neglect of intercellular spatial interactions and microenvironmental influencing factors. Single-cell independent analysis methods fail to capture biological interactions such as signal transduction, metabolic regulation, and functional synergy among cells, resulting in analysis results lacking microenvironmental contextual information. Fixed-threshold classification methods do not consider the differences in the microenvironments of different cells, assigning the same classification result to cells with the same staining intensity but different microenvironmental conditions. This simplistic approach significantly reduces the accuracy and biological significance of staining assessment. Existing data processing algorithms lack the ability to quantitatively describe cell population dynamics and tissue heterogeneity, failing to identify functional partitions and interaction patterns within tissues. Summary of the Invention

[0004] This application provides a data acquisition and analysis method and system for immunohistochemical staining machines, which solves the problem of lack of microenvironmental context information in existing immunohistochemical analysis by constructing a computational model based on intercellular signal transduction cascade reactions, thereby improving the biological accuracy of staining assessment.

[0005] In a first aspect, this application provides a method for data acquisition and analysis of an immunohistochemical staining machine, the method comprising:

[0006] The immunohistochemical sections were processed by multi-cell co-staining data acquisition and processing through neighborhood cell staining intensity correlation analysis to obtain an inter-cell staining correlation dataset.

[0007] Based on the spatial distribution pattern in the intercellular staining association dataset, the signal transduction intensity between neighboring cells is quantitatively modeled to obtain a signal transduction intensity matrix.

[0008] The signal transduction intensity matrix is ​​input into the cascade reaction prediction network to perform cascade calculations on the mutual influence mechanism of staining states between cells, and the microenvironment influence coefficient is obtained.

[0009] Based on the correction effect of the microenvironment influence coefficient on the staining state of the target cells, the single-cell staining results are subjected to context correction to obtain the corrected staining evaluation value.

[0010] Based on the corrected staining evaluation values, the cell population staining pattern is processed by microenvironment perception analysis to obtain staining analysis results based on cell interactions.

[0011] Optionally, the step of performing multi-cell co-staining data acquisition and processing on immunohistochemical sections through neighborhood cell staining intensity correlation analysis to obtain an inter-cell staining correlation dataset includes:

[0012] Each cell in the immunohistochemical section was subjected to multispectral staining intensity separation and detection to obtain a single-cell multichannel staining intensity vector.

[0013] Based on the DAB staining specific intensity value in the single-cell multi-channel staining intensity vector, the cell centroid coordinates are adaptively processed to calculate the neighborhood radius, thus obtaining the cell dynamic neighborhood range.

[0014] Based on the cell dynamic neighborhood range, the correlation of staining intensity between cells in the neighborhood is quantified to obtain the inter-cell staining synergy index.

[0015] The intercellular staining synergy index is standardized according to spatial distance weights to obtain an intercellular staining association dataset.

[0016] Optionally, the step of quantitatively modeling the signal transduction intensity between neighboring cells based on the spatial distribution pattern in the intercellular staining association dataset to obtain a signal transduction intensity matrix includes:

[0017] The distance decay function is calculated on the cell spatial coordinates in the intercellular staining association dataset to obtain the intercellular distance decay weights.

[0018] The directionality of the staining intensity gradient between cells is determined based on the distance attenuation weight to obtain the signal transduction direction vector;

[0019] Based on the signal transduction direction vector, the transduction coefficient of the staining influence intensity between neighboring cells is quantified to obtain the intercellular signal transduction value.

[0020] The intercellular signal transduction values ​​are arranged in a matrix according to the cell spatial location index to obtain the signal transduction intensity matrix.

[0021] Optionally, the step of inputting the signal transduction intensity matrix into a cascade reaction prediction network to perform cascade calculations on the interaction mechanism of cell staining states to obtain the microenvironment influence coefficient includes:

[0022] The signal transduction intensity matrix is ​​input into the cascaded reaction prediction network for multi-layer forward propagation calculation to obtain the primary influence response value between cells.

[0023] The primary impact response value is processed by cascade reaction path tracing calculation to obtain a multi-hop cell impact propagation sequence;

[0024] Based on the multi-hop cell influence propagation sequence, the intensity of the cascade reaction is calculated by cumulative superposition to obtain the cumulative cascade influence intensity.

[0025] The cumulative cascaded influence intensity is normalized and weighted to obtain the microenvironment influence coefficient.

[0026] Optionally, the cascade reaction path tracing calculation processing of the primary impact response value to obtain a multi-hop cell impact propagation sequence includes:

[0027] A graph traversal path search is performed on the cell node connection relationships in the primary influence response values ​​to obtain a set of inter-cell propagation paths;

[0028] Based on the path length parameters in the set of intercellular propagation paths, the multi-hop propagation distance is hierarchically graded to obtain a hierarchical propagation path structure.

[0029] Based on the hierarchical propagation path structure, the influence attenuation coefficient of each propagation path is calculated in descending order to obtain the path attenuation weight sequence;

[0030] The path attenuation weight sequence is serialized and arranged according to the propagation time sequence to obtain the multi-hop cell influence propagation sequence.

[0031] Optionally, the context correction processing of single-cell staining results based on the correction effect of the microenvironment influence coefficient on the staining state of the target cells to obtain a corrected staining evaluation value includes:

[0032] The microenvironment influence coefficient and the original staining intensity of the target cells are weighted and fused to obtain the microenvironment correction factor.

[0033] Based on the microenvironment correction factor, the threshold of cell staining classification boundary is redefined to obtain a personalized staining determination threshold.

[0034] The personalized staining determination threshold is applied to the staining status of the target cells for re-evaluation to obtain the corrected staining evaluation value.

[0035] Optionally, the step of performing microenvironment-sensing analysis on the cell population staining pattern based on the corrected staining evaluation value to obtain staining analysis results based on cell interactions includes:

[0036] The corrected staining evaluation values ​​are clustered and grouped according to the cell population distribution area to obtain cell population staining pattern partitions;

[0037] Based on the staining intensity distribution characteristics in the cell population staining pattern partitions, the interaction intensity between each partition is quantitatively analyzed to obtain the inter-population interaction intensity matrix.

[0038] The cell population staining heterogeneity is comprehensively evaluated based on the inter-population interaction strength matrix to obtain the microenvironment heterogeneity index.

[0039] The results of integrating the microenvironment heterogeneity index with the cell population staining pattern partitioning are processed to obtain staining analysis results based on cell interactions.

[0040] Secondly, this application provides an immunohistochemical staining machine data acquisition and analysis system, the immunohistochemical staining machine data acquisition and analysis system comprising:

[0041] The acquisition module is used to acquire and process multi-cell co-staining data of immunohistochemical sections through neighborhood cell staining intensity correlation analysis to obtain inter-cell staining correlation datasets.

[0042] The modeling module is used to perform quantitative modeling of the signal transduction intensity between neighboring cells based on the spatial distribution pattern in the intercellular staining association dataset, and obtain the signal transduction intensity matrix.

[0043] The calculation module is used to input the signal transduction intensity matrix into the cascade reaction prediction network, perform cascade calculation processing on the interaction mechanism of staining states between cells, and obtain the microenvironment influence coefficient.

[0044] The correction module is used to perform context correction processing on the single-cell staining results based on the correction effect of the microenvironment influence coefficient on the staining state of the target cells, and obtain the corrected staining evaluation value.

[0045] The analysis module is used to perform microenvironment-sensing analysis on the cell population staining pattern based on the corrected staining evaluation value, and obtain staining analysis results based on cell interactions.

[0046] Thirdly, an immunohistochemical staining machine data acquisition and analysis device is provided, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor invokes the instructions in the memory to cause the immunohistochemical staining machine data acquisition and analysis device to execute the above-described immunohistochemical staining machine data acquisition and analysis method.

[0047] Fourthly, a computer-readable storage medium is provided, wherein instructions are stored therein, which, when executed on a computer, cause the computer to perform the above-described immunohistochemical staining machine data acquisition and analysis method.

[0048] The technical solution provided in this application establishes a cell-cell collaborative staining data acquisition mechanism through the correlation analysis of neighboring cell staining intensity, breaking through the technical limitations of traditional single-cell independent analysis. This transforms immunohistochemical data acquisition from isolated point detection to correlated network detection, thereby obtaining a multidimensional dataset containing information on cell-cell interactions. Based on the quantitative modeling processing technology of the spatial distribution pattern of the cell-cell staining correlation dataset, a signal transduction intensity matrix is ​​constructed, solving the technical problem that existing technologies cannot quantitatively describe the degree of cell-cell interaction, transforming the biological interaction between cells from a qualitative description to a precise numerical expression. The cascade reaction prediction network technology, through multi-layer forward propagation computation, achieves deep modeling of complex cell-cell interaction mechanisms. Compared with traditional linear analysis methods, this technology can capture nonlinear interactions and cascade amplification effects between cells, significantly enhancing the biological realism of immunohistochemical analysis. The context correction processing technology for the microenvironment influence coefficient, through the correction effect analysis of the target cell staining state, achieves personalized staining assessment, overcoming the rigid limitations of fixed threshold classification methods, ensuring that the staining assessment of each cell accurately reflects its true expression state in a specific microenvironment. The microenvironment perception and analysis technology, through systematic analysis of cell population staining patterns, constructs a heterogeneity description framework at the tissue level, filling the gap in traditional techniques for analyzing overall tissue characteristics and providing more comprehensive and in-depth information support for clinical pathological diagnosis.

[0049] In specific applications of immunohistochemical staining data acquisition and analysis, the core feature of the cascade reaction prediction network algorithm lies in its multi-layer neural network architecture, which can learn and model multi-hop propagation paths between cells. This algorithmic feature enables the system to identify indirect cell interaction patterns that traditional methods cannot detect. Particularly in tumor microenvironment analysis, this algorithmic feature can reveal the complex regulatory relationships between tumor cells, immune cells, and stromal cells, providing more accurate and comprehensive technical support for biomarker evaluation in precision medicine. The path tracing algorithm for multi-hop cell influence propagation sequences, through graph traversal and hierarchical processing, can systematically identify and quantify multi-level transmission chains between cells. The unique value of this algorithmic feature lies in its ability to discover cascade effects and amplification mechanisms between cells, extending immunohistochemical analysis from static staining intensity assessment to dynamic signal propagation analysis. Attached Figure Description

[0050] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0051] Figure 1 This is a schematic diagram of one embodiment of the immunohistochemical staining machine data acquisition and analysis method in this application.

[0052] Figure 2 This is a schematic diagram of the cell population staining pattern zoning microenvironment perception analysis results in the embodiments of this application;

[0053] Figure 3 This is a schematic diagram of one embodiment of the immunohistochemical staining machine data acquisition and analysis system in this application.

[0054] Figure 4 This is a schematic block diagram of the structure of the immunohistochemical staining machine data acquisition and analysis device in an embodiment of the present invention. Detailed Implementation

[0055] This application provides a method and system for data acquisition and analysis of an immunohistochemical staining machine. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0056] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 One embodiment of the immunohistochemical staining machine data acquisition and analysis method in this application includes:

[0057] Step S101: Collect and process multi-cell co-staining data of immunohistochemical sections by performing neighborhood cell staining intensity correlation analysis to obtain inter-cell staining correlation dataset;

[0058] Step S102: Based on the spatial distribution pattern in the intercellular staining association dataset, perform quantitative modeling of the signal transduction intensity between neighboring cells to obtain the signal transduction intensity matrix;

[0059] Step S103: Input the signal transduction intensity matrix into the cascade reaction prediction network, perform cascade calculation on the interaction mechanism of staining states between cells, and obtain the microenvironment influence coefficient.

[0060] Step S104: Based on the correction effect of the microenvironment influence coefficient on the staining state of the target cells, perform context correction on the single-cell staining results to obtain the corrected staining evaluation value.

[0061] Step S105: Perform microenvironment perception analysis on the cell population staining pattern based on the corrected staining evaluation value to obtain staining analysis results based on cell interactions.

[0062] It is understood that the executing entity of this application can be an immunohistochemical staining machine data acquisition and analysis system, or it can be a terminal or a server; the specific implementation is not limited here. This application's embodiments use a server as an example for illustration.

[0063] Specifically, neighborhood cell staining intensity correlation analysis refers to the quantitative detection of DAB staining specific intensity of each cell using multispectral separation technology. The single-cell multi-channel staining intensity vector contains the intensity values ​​of the three RGB channels and the concentration information of the DAB-specific staining marker. Then, based on the DAB staining intensity value of the cell, its neighborhood radius is dynamically calculated. Cells with higher DAB staining intensity have larger neighborhood radii because highly expressed cells have a wider range of influence on the surrounding environment. Next, the synergy of staining intensity between cells is calculated within the determined neighborhood. The synergy is determined by comparing the similarity of staining patterns of each cell in the neighborhood. Cells with similar staining patterns have a higher synergy index. Finally, the synergy index is weighted and standardized according to spatial distance. Cells that are closer in distance have higher weights, thus forming an inter-cell staining correlation dataset. This dataset contains the correlation strength information between each cell and its neighboring cells.

[0064] Spatial distribution patterns refer to the two-dimensional spatial arrangement characteristics of cells in tissue sections, which are identified by analyzing the coordinate information in the intercellular staining association dataset. Quantitative modeling of signal transduction intensity first requires calculating the distance attenuation weight between cells. The attenuation weight decreases exponentially with increasing intercellular distance, simulating the propagation and attenuation law of biological signals in tissues. Then, based on the distance attenuation weight, the gradient change of staining intensity between cells is analyzed. The staining intensity gradient reflects the directionality of signal transduction, and the gradient direction from high to low intensity represents the dominant direction of signal transduction, forming a signal transduction direction vector. Next, the influence intensity of the direction vector on the neighboring cells is quantified by the transduction coefficient. The transduction coefficient comprehensively considers the distance factor and the direction factor. Cells that are close to each other and are in the dominant direction of signal transduction have a higher transduction coefficient. Finally, the signal transduction values ​​between cells are arranged into a matrix according to their spatial location index, forming a signal transduction intensity matrix. The rows and columns of this matrix correspond to different cells, and the matrix element values ​​represent the signal transduction intensity between the corresponding cells.

[0065] The cascade reaction prediction network is a deep neural network specifically designed to handle multi-level interactions between cells. Its architecture includes multiple hidden layers to capture cascade effects at different levels. After inputting the signal transduction intensity matrix into the network, it processes the data through multi-layer forward propagation to obtain the primary interaction response values ​​between cells. These primary interaction response values ​​reflect the first-order interaction strength between directly adjacent cells. Then, cascade reaction path tracing is performed on these primary interaction response values. Path tracing uses a graph traversal algorithm to identify multi-hop propagation paths between cells. Each path represents the sequence of intermediate cells through which the signal propagates from the source cell to the target cell. Based on the path... The length of the path is hierarchically graded. The first-hop path is a directly adjacent cell, the second-hop path requires passing through an intermediate cell, and so on, forming a hierarchical propagation path structure. The influence attenuation coefficient is calculated for each path. The longer the path, the greater the attenuation. The path attenuation weight sequence is obtained and then arranged according to the propagation time to form a multi-hop cell influence propagation sequence. Based on this sequence, the intensity of the cascade reaction is cumulatively calculated. The accumulation process considers all possible propagation paths and their weights to obtain the cumulative cascade influence intensity. Finally, the microenvironment influence coefficient is obtained through normalization. This coefficient quantifies the comprehensive influence of the microenvironment on the staining state of each cell.

[0066] Context correction refers to the process of revising staining assessment results based on the degree of influence of the cell's microenvironment. First, the microenvironment influence coefficient and the original staining intensity of the target cell are weighted and fused. This fusion process integrates microenvironment information into single-cell staining information through weighted averaging. The weighting coefficient is dynamically adjusted according to the significance of the microenvironment influence; the more significant the influence, the higher the weight, resulting in a microenvironment correction factor. This factor reflects the cell staining state after considering the microenvironment influence. Then, the staining classification boundary threshold is redefined based on the correction factor. Traditional fixed-threshold classification methods ignore microenvironment differences, while personalized staining judgment thresholds are dynamically adjusted according to the microenvironment characteristics of each cell. Cells in an inhibitory microenvironment will have their thresholds increased accordingly, while cells in a promoting microenvironment will have their thresholds appropriately decreased. Finally, the personalized judgment threshold is applied to the target cell for reassessment, yielding a corrected staining assessment value. This value more accurately reflects the true staining state of the cell in the tissue microenvironment than traditional single-cell analysis results.

[0067] The microenvironment perception analysis first clusters the corrected staining evaluation values ​​according to the cell population distribution area. The clustering algorithm divides tissue sections into different functional regions based on cell spatial location and staining characteristics. Cells within each region share similar staining patterns and microenvironmental characteristics, forming cell population staining pattern partitions. Then, the interaction strength between these partitions is analyzed. The degree of interaction between partitions is quantitatively assessed by calculating the signal transduction intensity and influence range between cells in different partitions, resulting in an inter-population interaction strength matrix. This matrix describes the mutual influence relationships between different cell populations. Based on the interaction strength matrix, a comprehensive evaluation of cell population staining heterogeneity is performed. This heterogeneity evaluation considers staining consistency within partitions and interaction differences between partitions; a higher heterogeneity index indicates greater staining differences between different regions within the tissue. Significantly, the higher the complexity of the microenvironment, the more effective the heterogeneity index becomes. Finally, the heterogeneity index is integrated with the cell population staining pattern partitioning to form a staining analysis result based on cell interactions. This result not only includes the corrected staining assessment of each cell, but also provides information on the overall microenvironmental heterogeneity of the tissue and the interaction patterns between cell populations. Compared with traditional single-cell independent analysis methods, this invention can identify synergistic and antagonistic effects between cells and discover staining patterns hidden in cell populations. For example, in lung cancer tissue analysis, the PD-L1 staining intensity of a certain tumor cell was originally at the weak positive boundary, but after considering the high expression state of the surrounding immune cells and the influence of signal transduction, the corrected staining assessment value of the cell was reclassified as positive. At the same time, it was found that an immunosuppressive microenvironment was formed in this area, and there was a significant mutual promotion effect between different cell populations.

[0068] In one specific embodiment, the process of performing step S101 may specifically include the following steps:

[0069] Each cell in the immunohistochemical section was subjected to multispectral staining intensity separation and detection to obtain a single-cell multichannel staining intensity vector.

[0070] Based on the DAB staining specific intensity value in the single-cell multi-channel staining intensity vector, the cell centroid coordinates are adaptively processed to calculate the neighborhood radius, thus obtaining the cell dynamic neighborhood range.

[0071] Based on the dynamic neighborhood range of cells, the correlation of staining intensity among cells in the neighborhood is quantified to obtain the intercellular staining synergy index.

[0072] The intercellular staining synergy index was standardized according to spatial distance weights to obtain the intercellular staining association dataset.

[0073] Specifically, multispectral staining intensity separation and detection refers to the technical process of spectrally separating and quantifying the staining signal of each cell in a tissue section using the optical detection system of an immunohistochemical staining machine. This process first uses a multi-wavelength light source to illuminate the stained section. Different wavelengths of light produce specific absorption and reflection with different staining reagents. The composite staining signal is decomposed into independent signals of each staining component by a spectral analyzer. The single-cell multi-channel staining intensity vector contains the optical density values ​​of the cell in the three basic channels of red, green and blue, as well as the specific intensity value of the DAB staining label. The DAB staining specific intensity value reflects the expression abundance of the target protein in the cell. The protein expression level is quantified by measuring the optical density of the DAB chromogenic product. The staining intensity vector of each cell is in the form of a four-dimensional array, corresponding to the red channel intensity, green channel intensity, blue channel intensity and DAB specific intensity, respectively. These values ​​are directly acquired and digitally stored by an optical density meter. The adaptive neighborhood radius calculation is an algorithm that dynamically determines the influence range of each cell based on the specific intensity value of DAB staining. The core logic of this algorithm is that the higher the DAB staining intensity of a cell, the higher the concentration of signaling molecules it secretes, and the larger the influence range on surrounding cells. The calculation process first extracts the coordinates of the cell centroid as the center point, and then determines the neighborhood radius based on the DAB intensity value of the cell through a linear mapping relationship. The mapping relationship converts the numerical range of DAB intensity values ​​into a spatial distance range. The neighborhood radius of cells with lower intensity values ​​is set to a smaller value, and the neighborhood radius of cells with higher intensity values ​​is increased accordingly. The dynamic neighborhood range of a cell is a circular area with the cell centroid as the center and the calculated radius value as the radius. All cells within this area are considered to be directly affected by the central cell.

[0074] Co-occurrence quantification is a data processing method that assesses intercellular functional coordination by comparing the similarity of staining patterns among different cells within a neighborhood. This process first identifies all neighboring cells within the dynamic neighborhood of each cell. Then, it calculates the correlation between the staining intensity vectors of the central cell and each neighboring cell. The correlation calculation uses a vector similarity algorithm, which determines the similarity by comparing the intensity distribution patterns of two cells across four staining channels. Cells with similar staining patterns have similar protein expression patterns and cellular functional states, indicating a synergistic relationship between them. A higher correlation value indicates a stronger synergy. The intercellular staining synergy index is derived by weighted averaging the correlation values ​​of all cell pairs within the neighborhood. The weighting coefficient is determined based on the spatial distance between cells; cells that are closer together have higher weights because spatially adjacent cells interact more directly and strongly. Association strength standardization is a data preprocessing step that normalizes and standardizes the intercellular staining synergy index according to spatial distance weights. This process first calculates the Euclidean distance between each pair of cells, and then converts the distance into a weight coefficient using a distance decay function. The distance decay function adopts an exponential decay model, and the weight decay is more obvious as the distance is greater. Next, the synergy index is multiplied by the corresponding distance weight to obtain a weighted synergy index. Finally, all weighted synergy indices are standardized to uniformly map the numerical range to the interval between zero and one. The standardized values ​​constitute the intercellular staining association dataset, which is stored in matrix form. The rows and columns of the matrix represent different cells, and the matrix element values ​​represent the standardized association strength between the corresponding cells.

[0075] In one specific embodiment, the process of performing step S102 may specifically include the following steps:

[0076] The distance decay function is calculated on the cell spatial coordinates in the intercellular staining association dataset to obtain the intercellular distance decay weights.

[0077] The directionality of the staining intensity gradient between cells is determined by the distance attenuation weight, resulting in a signal transduction direction vector.

[0078] The signal transduction direction vector is used to quantify the transduction coefficient of the staining influence intensity between neighboring cells to obtain the intercellular signal transduction value.

[0079] The intercellular signal transduction values ​​are arranged in a matrix according to the cell spatial location index to obtain the signal transduction intensity matrix.

[0080] Specifically, the distance decay function calculation is a mathematical processing method that simulates the attenuation law of biological signals propagating in tissues based on the spatial distance between cells. This processing first extracts the spatial coordinate information of each cell from the intercellular staining association dataset. The coordinate information includes the horizontal and vertical coordinates of the cell centroid in the tissue section. Then, the Euclidean distance between any two cells is calculated using the standard two-dimensional spatial distance formula. Next, the distance value is substituted into the exponential decay function for weight calculation. The exponential decay function simulates the physical law that the concentration of biological signal molecules decreases exponentially with increasing distance when they diffuse in tissues. The greater the distance between cells, the smaller the weight, indicating a weaker degree of mutual influence. The distance decay weight value ranges from zero to one. When the distance is zero, the weight is one, and when the distance is infinite, the weight approaches zero. This weight reflects the difference in the intensity of interaction between cells due to differences in spatial position. Directional transduction determination is an algorithm that determines the dominant direction of signal transduction by analyzing the staining intensity gradient between cells. This process uses a distance decay weight to weight the difference in staining intensity between cells. The staining intensity gradient refers to the rate and direction of change of staining intensity between adjacent cells. The gradient is determined by comparing the staining intensity difference between the central cell and its neighboring cells. The gradient direction points from cells with low staining intensity to cells with high staining intensity, representing the potential direction of signal transduction. The directional determination algorithm further considers the moderating effect of distance weight on the gradient. Cells that are closer to each other contribute more to the gradient calculation, while cells that are farther away contribute less. The signal transduction direction vector is obtained by normalizing the gradient direction. This vector contains the angular information of the transduction direction and the magnitude of the transduction intensity.

[0081] The transduction coefficient quantification process is a calculation procedure that numerically evaluates the actual influence intensity between cells based on the signal transduction direction vector. This process combines the signal transduction direction vector with the biological interaction pattern between cells. The calculation of the transduction coefficient considers three main factors: the consistency of the transduction direction, the influence of the transduction distance, and the magnitude of the transduction intensity. The consistency of the transduction direction is evaluated by comparing the degree of agreement between the predicted signal transduction direction and the actual observed direction of change in the staining pattern. The higher the consistency, the more accurate the predicted transduction direction, and the greater the transduction coefficient. The influence of the transduction distance is reflected by the distance attenuation weight obtained in the previous calculation. The greater the distance, the smaller the transduction coefficient. The magnitude of the transduction intensity is determined by the amplitude of the staining intensity gradient. The larger the gradient amplitude, the stronger the signal transduction, and the greater the transduction coefficient. The intercellular signal transduction value is obtained by weighted fusion of the above three factors. This value quantitatively represents the degree of influence of one cell on the staining state of another cell. Matrix arrangement processing is a data organization process that arranges intercellular signal transduction values ​​in an ordered manner according to the spatial location index of the cells and constructs a matrix structure. This process first assigns a unique location index to each cell in the tissue slice. The indexes are numbered according to the spatial distribution order of the cells in the slice, usually using a raster numbering method from left to right and from top to bottom. Then, the signal transduction values ​​calculated between each pair of cells are filled into the corresponding positions in the matrix. The row index of the matrix represents the signal-emitting cell, the column index represents the signal-receiving cell, and the matrix element value is the signal transduction value between the corresponding cells. The signal transduction intensity matrix is ​​a square matrix, and the dimension of the matrix is ​​equal to the total number of cells in the tissue slice. The diagonal elements of the matrix are usually set to zero because the transduction value of a cell to itself has no practical meaning. The off-diagonal elements reflect the mutual influence relationship between different cells.

[0082] In Ki67 immunohistochemical staining analysis of colorectal cancer tissue, two adjacent tumor cells were located at coordinates A and B, respectively. The intercellular distance was obtained by calculating the Euclidean distance between the two points. This distance was then substituted into an exponential decay function to calculate the distance decay weight, which reflects the degree of influence of spatial distance on intercellular interaction. Next, the difference in Ki67 staining intensity between the two cells was analyzed. Cell A had lower staining intensity, while cell B had higher staining intensity. The staining intensity gradient was obtained by calculating the intensity difference and its direction. The gradient direction pointed from cell A to cell B, representing the dominant direction of signal transduction. The gradient was adjusted using the distance decay weight to obtain the signal transduction direction vector. The quantity includes comprehensive information on the direction and intensity of conduction. Based on this direction vector, the conduction coefficient is further calculated. The conduction coefficient comprehensively considers three factors: directional consistency, distance influence, and intensity magnitude, to obtain the signal conduction value of cell A to cell B. This value quantitatively represents the degree of influence of cell A's proliferation state on cell B's proliferation activity. The signal conduction values ​​between all cells are filled into the matrix structure according to the spatial location index of the cells to form a signal conduction intensity matrix. This matrix not only records the intensity of direct interaction between each pair of cells, but also reflects the complex signal conduction network relationship between cells in the entire tumor tissue, solving the technical problem of neglecting the spatial interaction between cells in traditional immunohistochemical analysis.

[0083] In one specific embodiment, the process of performing step S103 may specifically include the following steps:

[0084] The signal transduction intensity matrix is ​​input into the cascaded response prediction network for multi-layer forward propagation calculation to obtain the primary influence response value between cells.

[0085] The primary impact response values ​​were processed by cascade reaction path tracing calculations to obtain multi-hop cell impact propagation sequences.

[0086] The cumulative cascade effect intensity is obtained by cumulative superposition calculation of the cascade effect intensity based on the multi-hop cell effect propagation sequence.

[0087] The cumulative cascading influence intensity is normalized and weighted to obtain the microenvironment influence coefficient.

[0088] Specifically, the cascade response prediction network is a deep neural network architecture specifically designed to handle multi-level interactions between cells. This network comprises an input layer, multiple hidden layers, and an output layer, specifically designed to capture cascade effects between cells. Multi-layer forward propagation computation refers to the process of using a signal transduction intensity matrix as input data and performing layer-by-layer computation through the neural network. The input layer receives the values ​​from the signal transduction intensity matrix, where each matrix element represents the direct signal transduction intensity between a pair of cells. The input layer converts the matrix data into a vector format required for network computation. The first hidden layer performs linear transformation and non-linear activation on the input vector. The linear transformation is achieved through matrix multiplication of the weight matrix and the input vector, while the non-linear activation function introduces non-linear features, enabling the network to learn complex cell interaction patterns. The second hidden layer continues to process the output of the first layer in a similar manner, extracting higher-level feature information layer by layer. Finally, the output layer generates the primary effect response value between cells. This response value reflects the strength of the direct interaction between cells after considering the network's learning. Compared to the original signal transduction intensity, the primary effect response value incorporates more contextual information and non-linear relationships. The cascade reaction path tracing computation is based on an algorithm that identifies and analyzes multi-hop propagation paths between cells using primary influence response values. This process models the interactions between cells as a directed graph structure, with each cell as a node in the graph and the primary influence response value as the edge weight between nodes. The path tracing algorithm uses graph traversal techniques to systematically search for all possible propagation paths from any source cell to the target cell. A one-hop path indicates that the source cell directly affects the target cell, a two-hop path indicates that the source cell affects the target cell through an intermediate cell, and so on to form multi-hop propagation paths. During the path search process, the algorithm records the cell sequence traversed by each path and the corresponding influence intensity value. The path length limit is set according to biological rationale, usually not exceeding five hops, because excessively long propagation paths have a weak impact on biological systems. The multi-hop cell influence propagation sequence contains the complete propagation chain from the source cell to the target cell, and each sequence records the identifier and propagation intensity of each cell on the propagation path.

[0089] Cumulative superposition calculation is a data processing method that comprehensively calculates the influence intensity of multiple propagation paths. This process first calculates the total influence intensity of each propagation path, obtained by multiplying or weighted averaging the propagation intensities of each segment along the path. The multiplication operation simulates the gradual attenuation of the signal during propagation, while the weighted averaging operation considers the relative importance of different propagation segments. Then, the intensities of all paths between the same pair of source and target cells are superimposed. Superposition methods include direct summation or weighted summation, with weighting coefficients determined based on path length; shorter paths have higher weights due to their higher propagation efficiency. The cumulative cascaded influence intensity comprehensively reflects the overall influence of the source cell on the target cell. This value considers both direct and indirect influences through intermediate cells. Compared to simple direct influence analysis, the cumulative influence intensity more comprehensively describes the complex interactions between cells. Normalized weighting is a data standardization step that converts the cumulative cascade effect intensity into standardized microenvironmental influence coefficients. This process first calculates the statistical characteristics of the cumulative effect intensity among all cells, including the maximum, minimum, mean, and standard deviation. Then, the min-max normalization method is used to map the influence intensity values ​​to a standard interval of zero to one. The normalization formula subtracts the minimum value from the original value and divides it by the numerical range to ensure that all coefficients are compared on the same scale. The weighting process is adjusted according to the relative position and functional importance of each cell in the tissue. Cells in the central region usually have a higher weight than cells in the peripheral region because central cells are influenced by more neighboring cells. The microenvironmental influence coefficients ultimately form a standardized numerical matrix, where the matrix elements represent the degree of comprehensive influence of the microenvironment on the corresponding cells.

[0090] In HER2 immunohistochemical staining analysis of gastric cancer tissue, the signal transduction intensity matrix contained all direct transduction relationships between cells. After inputting this matrix into a cascade reaction prediction network, the network identified complex interaction patterns between cells through multi-layer computation. The output primary impact response values ​​not only retained direct interaction information but also incorporated indirect correlations learned by the network. Based on these response values, path tracing revealed that a tumor cell with high HER2 expression not only directly affected its neighboring tumor cells but also indirectly affected cell populations at greater distances through intermediate stromal cells. The path tracing algorithm identified multiple propagation paths, including two-hop and three-hop paths. Each path recorded the cell propagation sequence and the corresponding impact intensity. Cumulative superposition calculations combined the impact intensities of all paths, revealing that the cumulative impact intensity of the high-expressing cell on distant cell populations was significantly higher than that of simple direct impact, indicating a significant cascade amplification effect between cells. Finally, normalization was used to convert the cumulative impact intensity into a microenvironmental impact coefficient, which quantitatively describes the degree of comprehensive influence of the surrounding microenvironment on each cell, solving the technical problem of the inability to quantify cascade interactions between cells in traditional analysis.

[0091] In one specific embodiment, the process of performing cascade reaction path tracing calculations on the primary impact response values ​​may specifically include the following steps:

[0092] A graph traversal path search is performed on the cell node connection relationships in the primary influence response values ​​to obtain a set of inter-cell propagation paths.

[0093] Based on the path length parameter in the set of intercellular propagation paths, the multi-hop propagation distance is hierarchically graded to obtain a hierarchical propagation path structure.

[0094] The path attenuation weight sequence is obtained by calculating the attenuation coefficient of each propagation path in descending order based on the hierarchical propagation path structure.

[0095] The path decay weight sequence is serialized and arranged according to the propagation time sequence to obtain the multi-hop cell influence propagation sequence.

[0096] Specifically, the graph traversal path search processing is an algorithm that constructs a cell connection graph based on primary influence response values ​​and systematically searches for all possible propagation paths. This processing converts the primary influence response values ​​between cells into a directed graph structure, with each cell as a node in the graph and the primary influence response value as the weight of the directed edge connecting the nodes. The connection relationship between cell nodes is determined by threshold filtering. When the influence response values ​​between two cells exceed a preset threshold, a connection is established. The graph traversal algorithm adopts a strategy combining depth-first search and breadth-first search. Starting from each source cell node, it recursively visits all reachable target cell nodes. During the search process, the algorithm records each propagation path, which includes the starting cell, intermediate cell sequence, and ending cell. The search depth is limited to a maximum of five levels to avoid overly complex long-distance propagation paths. The set of inter-cell propagation paths contains all possible propagation sequences from any source cell to any target cell. Each path records the cell identifiers passed through and the total path length. The data structure of the path set is stored in the form of an adjacency list, which facilitates subsequent path analysis and processing operations. Hierarchical tiered processing is a data structuring method that classifies and organizes a set of paths based on the hop count of the propagation path. This process first calculates the length parameter of each propagation path, which is equal to the number of cell nodes in the path minus one, i.e., the propagation hop count. Then, all paths are assigned to different hierarchical levels based on their length. One-hop paths are classified as level one, representing propagation between directly adjacent cells; two-hop paths are classified as level two, representing indirect propagation through an intermediate cell; and so on until the highest level. The hierarchical propagation path structure is stored using a multi-layered nested data structure. Each level contains all propagation paths of that length. Paths within a level are sorted from highest to lowest influence intensity, while paths between levels are sorted from shortest to longest. This structure facilitates subsequent hierarchical path analysis and weight calculation.

[0097] The decreasing calculation processing is an algorithm that calculates the attenuation weight for each path based on the principle of gradual attenuation of biological signals during multi-hop propagation. This processing calculates the influence attenuation coefficient for each path in the hierarchical propagation path structure. The calculation of the attenuation coefficient takes into account the influence of path length on signal propagation effect. The longer the path, the more severe the signal attenuation. The attenuation calculation adopts an exponential attenuation model, and the attenuation coefficient is equal to the attenuation constant raised to the power of the path length. The attenuation constant is set according to biological experience, usually between 0.5 and 0.8. The attenuation coefficients of different path lengths are significantly different. The attenuation coefficient of the one-hop path is the highest, the attenuation coefficient of the two-hop path is about half to three-quarters of that of the one-hop path, and the attenuation coefficient of the three-hop path decreases further. The path attenuation weight sequence records the attenuation coefficient corresponding to each propagation path. The weight values ​​in the sequence reflect the degree of contribution of different paths to the final influence effect. The calculation of attenuation weight also considers the product effect of the propagation intensity of each segment on the path. The weaker propagation segment will further reduce the effective weight of the entire path. Serialization is a data organization method that reorganizes the path attenuation weight sequence according to the chronological order of signal propagation. This process first determines the propagation time sequence of each propagation path, which is based on the path length and propagation direction. Paths of equal length are sorted according to the priority of the starting cell, which is determined by a combination of cell staining intensity and spatial location. Cells with high staining intensity and located in the central region have higher priority. Then, all paths are arranged according to the propagation time sequence to form an ordered propagation sequence. Each element in the sequence contains a path identifier, propagating cell sequence, attenuation weight, and temporal position. The multi-hop cell influence propagation sequence reflects the complete spatiotemporal dynamics of signal propagation between cells. This sequence not only contains detailed information about each propagation path but also reflects the relative importance and propagation order between different paths.

[0098] In one specific embodiment, the process of executing step S104 may specifically include the following steps:

[0099] The microenvironment influence coefficient and the original staining intensity of the target cells were weighted and fused to obtain the microenvironment correction factor.

[0100] Based on the microenvironment correction factor, the threshold of cell staining classification boundary is redefined to obtain a personalized staining judgment threshold.

[0101] The personalized staining judgment threshold is applied to the staining status of the target cells for re-evaluation, resulting in a corrected staining evaluation value.

[0102] Specifically, the weighted fusion calculation is a mathematical operation that numerically integrates the microenvironmental influence coefficient with the original staining intensity of the target cell. This process first extracts the original staining intensity value of the target cell, which is directly derived from the optical density detection results of the immunohistochemical staining machine, reflecting the expression abundance of the target protein within the cell. Then, the corresponding microenvironmental influence coefficient is obtained. This coefficient is calculated through the aforementioned cascade reaction analysis, quantifying the comprehensive influence of surrounding cells on the target cell. The weighted fusion uses a linear weighted average method, multiplying the original staining intensity by the original weight coefficient and the microenvironmental influence coefficient by the environmental weight coefficient. The two weighted results are then added to obtain the fused value. The allocation of weight coefficients is determined according to the significance of the microenvironmental influence; when the microenvironmental influence coefficient is high, the environmental weight increases; when the microenvironmental influence is weak, the original weight dominates. The sum of the weight coefficients always equals one to ensure the numerical stability of the fusion result. The microenvironmental correction factor comprehensively reflects the actual staining state of the cell after considering the microenvironmental influence. This factor more accurately describes the true expression level of the cell in the tissue microenvironment than the original staining intensity. The threshold redefinition process is an algorithm that personalizes the traditional fixed staining classification boundary based on microenvironment correction factors. This process first analyzes the distribution characteristics of the microenvironment correction factors and calculates statistical parameters such as the mean, variance, and quantiles of these factors. Traditional staining classification uses a fixed threshold to classify cells into three categories: negative, weakly positive, and strongly positive. However, the fixed threshold ignores the regulatory effect of different microenvironments on cell expression. Personalized threshold adjustment dynamically sets the threshold based on the microenvironment characteristics of each cell. When cells are in a promoting microenvironment, the classification threshold is appropriately lowered, making it easier for cells that are originally at the boundary to be classified as positive. When cells are in an inhibiting microenvironment, the classification threshold is correspondingly raised to avoid misclassifying inhibited cells as strongly positive. The personalized staining determination threshold is calculated by analyzing the deviation between the correction factors and the traditional classification results. Cells with larger deviations require more significant threshold adjustments, while cells with smaller deviations require relatively smaller threshold adjustments.

[0103] The reassessment process applies personalized staining thresholds to target cells for reclassification. This process compares each cell's microenvironment correction factor with its corresponding personalized threshold. When the correction factor exceeds the personalized threshold, the cell is reclassified to a higher staining grade; when the correction factor falls below the personalized threshold, the cell is reclassified to a lower staining grade. The reclassification process employs a multi-level threshold system, including thresholds for negative and weakly positive staining, and weakly positive and strongly positive staining. Each threshold is individually adjusted based on the cell's microenvironment characteristics. The classification result includes not only the staining grade but also a confidence score. The confidence score is determined by the difference between the correction factor and the threshold; a larger difference indicates higher classification confidence. The corrected staining assessment value is the final classification result after reassessment. This value combines the reclassified staining grade and the corresponding confidence score, forming a more accurate description of staining status than traditional single-cell analysis.

[0104] In one specific embodiment, the process of executing step S105 may specifically include the following steps:

[0105] The corrected staining evaluation values ​​were clustered and grouped according to the cell population distribution area to obtain cell population staining pattern partitions;

[0106] Based on the staining intensity distribution characteristics in the cell population staining pattern partitions, the interaction intensity between each partition was quantitatively analyzed to obtain the interpopulation interaction intensity matrix.

[0107] The cell population staining heterogeneity was comprehensively evaluated based on the inter-population interaction strength matrix to obtain the microenvironment heterogeneity index.

[0108] The results of the microenvironment heterogeneity index and cell population staining pattern partitioning were integrated to obtain staining analysis results based on cell interactions.

[0109] Specifically, clustering grouping is a data mining method that classifies cells in tissue sections based on spatial and functional similarity using corrected staining evaluation values. This process first extracts the corrected staining evaluation value for each cell as a clustering feature. The corrected staining evaluation value comprehensively reflects the staining intensity and microenvironmental influence of the cell. Then, it combines the spatial coordinate information of the cells to construct a multi-dimensional feature vector, which includes three dimensions: staining intensity, spatial location, and microenvironmental coefficient. The clustering algorithm uses the K-means clustering method, iteratively calculating to group cells with similar characteristics into the same category. During the clustering process, the algorithm automatically determines the optimal number of clusters. The clustering effect is evaluated by the silhouette coefficient and the ratio of intra- and inter-cluster distances. The number of clusters is typically set between three and eight, corresponding to different cell functional groups. The cell population distribution area is spatially divided on the tissue section according to the clustering results. Each region contains cell populations with similar staining patterns. The cell population staining pattern partitioning reflects the functional heterogeneity distribution within the tissue. Different partitions represent cell populations with different biological characteristics. The partition boundaries are determined by dual constraints of spatial continuity and staining similarity. Quantitative analysis is a numerical evaluation method for calculating the interaction strength between different cell populations. This process first calculates the staining intensity distribution characteristics of each partition, including four statistical parameters: mean, variance, skewness, and kurtosis. The mean reflects the average expression level of cells within the partition, the variance reflects the uniformity of expression, the skewness describes the asymmetry of the distribution, and the kurtosis describes the sharpness of the distribution. Then, the differences in staining intensity distribution between different partitions are analyzed. These differences are quantified by distance measures between distributions, including differences in mean, distribution shape, and overlap. Next, based on the aforementioned signal transduction analysis results, the interaction strength between partitions is calculated. The interaction strength is determined by statistically analyzing the number and intensity of cross-partition signal transduction paths. The more transduction paths and the greater the transduction intensity between partitions, the higher the interaction strength. The inter-population interaction strength matrix records the degree of interaction between any two partitions, with rows and columns corresponding to different partitions, and matrix element values ​​representing the interaction strength between the corresponding partitions.

[0110] The comprehensive evaluation process is an analytical method that quantitatively describes the complexity and heterogeneity of the entire organizational microenvironment based on the interaction strength matrix between groups. This process first calculates the global characteristic parameters of the interaction matrix, including linear algebraic features such as the spectral radius, condition number, and rank. The spectral radius reflects the strongest interaction intensity, the condition number reflects the stability of the interaction, and the matrix rank reflects the number of independent interaction patterns. Next, the degree of non-uniformity of the matrix is ​​analyzed. Non-uniformity is assessed by calculating the standard deviation and coefficient of variation of the matrix elements. A larger standard deviation indicates more significant differences in interactions between different partitions, and the coefficient of variation reflects the degree of relative difference. Then, the network connectivity index is calculated. The interaction matrix is ​​treated as a weighted undirected graph, and network topology parameters such as connectivity, clustering coefficient, and average path length are calculated. Connectivity reflects the density of interactions between partitions, the clustering coefficient reflects the tightness of local interactions, and the average path length reflects the efficiency of information propagation. The microenvironmental heterogeneity index is obtained by weighted comprehensive calculation of the above parameters. The weight coefficients are determined according to the importance of each parameter's contribution to heterogeneity. A higher heterogeneity index indicates a more complex and heterogeneous organizational microenvironment. The results integration processing is a data synthesis method that maps the microenvironmental heterogeneity index with cell population staining pattern partitions. This processing uses the heterogeneity index as a whole tissue characteristic parameter and performs correlation analysis with the local features of each partition. The correlation analysis identifies which partitions contribute the most to the overall heterogeneity and which partitions have the most significant interactions. Then, a multi-level analysis result structure is constructed, which includes three levels: the overall tissue level, the partition population level, and the single-cell level. The overall level provides the heterogeneity index and overall staining pattern description, the population level provides the staining characteristics and interaction relationships of each partition, and the single-cell level provides the corrected staining assessment value and microenvironmental impact information for each cell. The staining analysis results based on cell interactions integrate traditional single-cell staining information and innovative intercellular interaction information to form a tissue microenvironment analysis report.

[0111] like Figure 2As shown in the figure, this embodiment divides tissue sections into four cell population partitions—high expression, medium expression, low expression, and mixed expression—after clustering the corrected staining evaluation values. The black square lines in the figure represent the microenvironmental heterogeneity index of each partition, while the gray triangle lines represent the interaction strength between populations. The biaxial display results show that the high expression and mixed expression regions have higher microenvironmental heterogeneity indices, at 0.72 and 0.85 respectively, while the low expression region has the lowest heterogeneity index at 0.28. The interaction strength between populations shows a similar trend to the heterogeneity index, with the mixed expression region reaching the highest value of 0.79. This indicates a complex regulatory relationship between different cell types within this region. This analysis effectively reveals the functional differentiation pattern and dynamic intercellular interaction network within the immunohistochemically stained tissue, providing a quantitative description of microenvironmental characteristics for cell-interaction-based staining analysis.

[0112] The data acquisition and analysis method of the immunohistochemical staining machine in the embodiments of this application has been described above. The data acquisition and analysis system of the immunohistochemical staining machine in the embodiments of this application is described below. Please refer to [link to relevant documentation]. Figure 3 One embodiment of the immunohistochemical staining machine data acquisition and analysis system in this application includes:

[0113] The acquisition module is used to acquire and process multi-cell co-staining data of immunohistochemical sections through neighborhood cell staining intensity correlation analysis to obtain inter-cell staining correlation datasets.

[0114] The modeling module is used to perform quantitative modeling of the signal transduction intensity between neighboring cells based on the spatial distribution pattern in the intercellular staining association dataset, and obtain the signal transduction intensity matrix.

[0115] The calculation module is used to input the signal transduction intensity matrix into the cascade reaction prediction network, perform cascade calculation processing on the interaction mechanism of staining states between cells, and obtain the microenvironment influence coefficient.

[0116] The correction module is used to perform context correction processing on the single-cell staining results based on the correction effect of the microenvironment influence coefficient on the staining state of the target cells, and obtain the corrected staining evaluation value.

[0117] The analysis module is used to perform microenvironment-sensing analysis on the cell population staining pattern based on the corrected staining evaluation value, and obtain staining analysis results based on cell interactions.

[0118] above Figure 3 The immunohistochemistry staining machine data acquisition and analysis system in this embodiment of the invention is described in detail from the perspective of modular functional entities. The immunohistochemistry staining machine data acquisition and analysis device in this embodiment of the invention is described in detail from the perspective of hardware processing.

[0119] Reference Figure 4 This invention also provides an immunohistochemical staining machine data acquisition and analysis device, which can be a server, and its internal structure can be as follows: Figure 4 As shown, the immunohistochemistry staining machine data acquisition and analysis device includes a processor, memory, display screen, input device, network interface, and database connected via a system bus. The processor in this computer design provides computing and control capabilities. The memory of the immunohistochemistry staining machine data acquisition and analysis device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database of the immunohistochemistry staining machine data acquisition and analysis device is used to store the data corresponding to this embodiment. The network interface of the immunohistochemistry staining machine data acquisition and analysis device is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements the above-described method.

[0120] Those skilled in the art will understand that Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present invention and does not constitute a limitation on the immunohistochemical staining machine data acquisition and analysis equipment to which the present invention is applied.

[0121] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a computer, cause the computer to perform the steps of the immunohistochemical staining machine data acquisition and analysis method.

[0122] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0123] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, 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. This computer software product is stored in a storage medium and includes several instructions to cause an immunohistochemical staining machine data acquisition and analysis device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0124] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for data acquisition and analysis using an immunohistochemical staining machine, characterized in that, include: Each cell in the immunohistochemical section was subjected to multispectral staining intensity separation and detection to obtain a single-cell multichannel staining intensity vector. Based on the DAB staining-specific intensity value in the single-cell multi-channel staining intensity vector, the cell centroid coordinates are adaptively processed to calculate the neighborhood radius, thus obtaining the cell dynamic neighborhood range; based on the cell dynamic neighborhood range, the correlation of staining intensity between cells within the neighborhood is quantified to obtain the inter-cell staining synergy index. The intercellular staining synergy index is standardized according to spatial distance weights to obtain an intercellular staining association dataset. The association strength standardization process includes: calculating the Euclidean distance between each pair of cells, converting the distance into a weight coefficient using a distance decay function, multiplying the synergy index by the corresponding distance weight to obtain a weighted synergy index, standardizing all weighted synergy indices, and mapping the numerical range uniformly to the interval between zero and one. The standardized values ​​constitute the intercellular staining association dataset. Based on the spatial distribution pattern in the intercellular staining association dataset, the signal transduction intensity between neighboring cells is quantitatively modeled to obtain the signal transduction intensity matrix. The signal transduction intensity matrix is ​​input into the cascade reaction prediction network to perform cascade calculations on the interaction mechanism of staining states between cells, and the microenvironment influence coefficient is obtained. The microenvironment influence coefficient and the original staining intensity of the target cells are weighted and fused to obtain the microenvironment correction factor; the cell staining classification boundary is redefined based on the microenvironment correction factor to obtain the personalized staining judgment threshold; the personalized staining judgment threshold is applied to the staining status of the target cells for re-evaluation to obtain the corrected staining evaluation value. The corrected staining evaluation values ​​were clustered and grouped according to the cell population distribution area to obtain cell population staining pattern partitions. Based on the staining intensity distribution characteristics in the cell population staining pattern partitions, the interaction intensity between each partition was quantitatively analyzed to obtain the inter-population interaction intensity matrix. Based on the inter-population interaction intensity matrix, the cell population staining heterogeneity was comprehensively evaluated to obtain the microenvironment heterogeneity index. The microenvironment heterogeneity index and the cell population staining pattern partitions were integrated to obtain the staining analysis results based on cell interactions.

2. The data acquisition and analysis method for immunohistochemical staining machine according to claim 1, characterized in that, Based on the spatial distribution pattern in the intercellular staining association dataset, the signal transduction intensity between neighboring cells is quantitatively modeled to obtain a signal transduction intensity matrix. This includes: calculating the distance attenuation function on the spatial coordinates of cells in the intercellular staining association dataset to obtain intercellular distance attenuation weights; determining the directional transmission of the intercellular staining intensity gradient based on the distance attenuation weights to obtain a signal transduction direction vector; quantifying the transmission coefficient of the influence intensity of staining between neighboring cells based on the signal transduction direction vector to obtain intercellular signal transduction values; and arranging the intercellular signal transduction values ​​in a matrix according to the cell spatial location index to obtain the signal transduction intensity matrix.

3. The data acquisition and analysis method for immunohistochemical staining machine according to claim 1, characterized in that, The signal transduction intensity matrix is ​​input into the cascade reaction prediction network to perform cascade calculations on the interaction mechanism of cell staining states, resulting in the microenvironment influence coefficient. This process includes: inputting the signal transduction intensity matrix into the cascade reaction prediction network for multi-layer forward propagation calculations to obtain the primary influence response value between cells; performing cascade reaction path tracing calculations on the primary influence response value to obtain the multi-hop cell influence propagation sequence; performing cumulative superposition calculations on the cascade reaction intensity based on the multi-hop cell influence propagation sequence to obtain the cumulative cascade influence intensity; and performing normalized weight allocation on the cumulative cascade influence intensity to obtain the microenvironment influence coefficient.

4. The data acquisition and analysis method for immunohistochemical staining machine according to claim 3, characterized in that, The primary impact response values ​​are processed by cascade reaction path tracing calculation to obtain a multi-hop cell impact propagation sequence. This includes: performing graph traversal path search on the cell node connection relationships in the primary impact response values ​​to obtain a set of inter-cell propagation paths; performing hierarchical classification of the multi-hop propagation distance based on the path length parameters in the inter-cell propagation path set to obtain a hierarchical propagation path structure; calculating the decreasing impact attenuation coefficient of each propagation path based on the hierarchical propagation path structure to obtain a path attenuation weight sequence; and arranging the path attenuation weight sequence according to the propagation time sequence to obtain the multi-hop cell impact propagation sequence.

5. A data acquisition and analysis system for an immunohistochemical staining machine, characterized in that, For implementing the immunohistochemical staining machine data acquisition and analysis method as described in any one of claims 1-4, the immunohistochemical staining machine data acquisition and analysis system comprises: The acquisition module is used to acquire and process multi-cell co-staining data of immunohistochemical sections through neighborhood cell staining intensity correlation analysis to obtain inter-cell staining correlation datasets. The modeling module is used to perform quantitative modeling of the signal transduction intensity between neighboring cells based on the spatial distribution pattern in the intercellular staining association dataset, and obtain the signal transduction intensity matrix. The calculation module is used to input the signal transduction intensity matrix into the cascade reaction prediction network, perform cascade calculations on the interaction mechanism of staining states between cells, and obtain the microenvironment influence coefficient. The correction module is used to correct the staining state of target cells based on the microenvironment influence coefficient, and to perform context correction on single-cell staining results to obtain corrected staining evaluation values. The analysis module is used to perform microenvironment-sensing analysis on cell population staining patterns based on corrected staining assessment values, and obtain staining analysis results based on cell interactions.

6. A data acquisition and analysis device for an immunohistochemical staining machine, characterized in that, It includes a memory and a processor, the memory storing a computer program that can run on the processor, and the processor executing the computer program to implement the immunohistochemical staining machine data acquisition and analysis method according to any one of claims 1 to 5.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When a computer program is run by a processor, it causes the processor to execute the immunohistochemical staining machine data acquisition and analysis method as described in any one of claims 1 to 5.

Images