Method, system, device and medium for quantitative analysis of cancerous regions of tissue sections

Abstract

The application discloses a cancerous region quantitative analysis method, system, device and medium of tissue section, the method comprises: obtaining the continuous section and the staining diagram of H&E section of the same piece of cancerous tissue; and extracting the cell nucleus coordinate and the biomarker fluorescence intensity of each staining diagram from the staining diagram; through two-stage ICP registration, namely, coarse registration of the original coordinate system and fine registration of the normalized coordinate system, realizing three-section coordinate unification with H&E section as the reference, screening the reference section and transforming the target section; selecting the same cancerous region after registration for quantitative statistical analysis of the deviation coefficient of the positive rate. The application adopts the cell nucleus coordinate point cloud as the core data in the ICP registration, avoids the interference of the staining difference and the imaging noise on the registration reference from the source, realizes the global alignment of the tissue scale and the single-cell-level spatial consistency, improves the accuracy and repeatability of the cross-section pathological analysis, and is suitable for the spatial pattern research of cancer and tumor microenvironment.

Description

Technical Field

[0001] This application belongs to the field of statistical analysis technology, and in particular relates to a method, system, device and medium for quantitative analysis of cancerous areas in tissue sections. Background Technology

[0002] In pathological research and translational applications, spatial registration and quantitative analysis between multiple sections / modalities (such as H&E and multi-marker fluorescent sections) are fundamental to reproducing the spatial patterns of cancer and the tumor microenvironment, and to conducting cross-section statistical and comparative assessments. An ideal technical approach should achieve global alignment across sections at the tissue scale, enabling successful mapping of cancer regions from H&E sections back to the two sections requiring consistency checks, while maintaining spatial consistency at the single-cell scale. This ensures the consistency and reproducibility of subsequent gridded statistical analysis and positive result determination.

[0003] Most mainstream methods are based on image-level registration and statistics. For example, HALO's "Serial Section Analysis Add-on" module performs registration based on image structure and staining information (bright field or fluorescence), and users can choose to use a stable channel (e.g., DAPI kernel staining) for alignment. Intensity-based registration, on the other hand, achieves alignment by directly optimizing the pixel (or voxel) intensity similarity of the entire image. It often uses rigid, affine, or similar transformations, as well as more flexible non-rigid transformations (such as B-spline free deformation field FFD, Demons algorithm, SyN, or LDDMM isopreserving differential deformation), and optimizes stepwise based on a multi-resolution pyramid strategy, combined with regularization to control smoothness. It is applicable to both monomodal and multimodal images. Feature / keypoint-driven registration methods first extract local invariant features (such as SIFT, SURF, ORB) or geometric structures such as corners and edges, estimate the initial transformation matrix through RANSAC or random sampling consensus algorithms, and then perform global or local fine-tuning.

[0004] Currently, these mainstream image-level registration methods are easily affected by staining differences and imaging conditions. Pixels, textures, and feature points can vary significantly due to staining depth, exposure, noise, and batch effects, leading to unstable feature matching and a tendency for local drift or deformation overfitting. When analyzing cell states in specific regions, image-level registration alone often only works on pixel or image patch units, making it difficult to directly constrain the correspondence of "nuclear coordinates" across slices. This results in insufficient spatial consistency at the single-cell level, affecting the accuracy and reproducibility of cell-level statistics and result interpretation. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a method, system, device and medium for quantitative analysis of cancerous areas in tissue sections, so as to solve the problems of insufficient cross-section registration accuracy, poor spatial consistency at the single-cell level and unreliable quantification of cancerous areas.

[0006] To address the aforementioned technical problems, this application provides the following technical solution:

[0007] Firstly, this application provides a method for quantitative analysis of cancerous regions in tissue sections, including:

[0008] Obtain the staining images corresponding to the first slice, the second slice, and the H&E slice, wherein the first slice and the second slice are two consecutive tissue slices from the same cancerous tissue, the staining image of the first slice is denoted as the first staining image, the staining image of the second slice is denoted as the second staining image, and the staining image of the H&E slice is denoted as the third staining image.

[0009] Image information is extracted from the first staining image, the second staining image, and the third staining image to obtain the information extraction results corresponding to each staining image. The information extraction results of the first staining image and the second staining image include: data of cell nuclear coordinates and fluorescence intensity of biomarkers in the staining image. The information extraction results of the third staining image include: data of cell nuclear coordinates, fluorescence intensity of biomarkers, and information of cancerous areas in the staining image.

[0010] Based on the information extraction results corresponding to each coloring image, ICP registration processing is performed on the first coloring image, the second coloring image, and the third coloring image to obtain the alignment matrix after registration processing.

[0011] Based on the alignment matrix, a quantitative statistical analysis is performed on the deviation coefficient of the positive rate of the same cancerous region in the first slice and the second slice.

[0012] Further, the step of extracting image information from the first, second, and third stained images to obtain corresponding information extraction results includes:

[0013] The first, second, and third stained images were analyzed using StrataQuest software to extract image information and obtain the corresponding information extraction results.

[0014] Further, the step of performing ICP registration processing on the first coloring image, the second coloring image, and the third coloring image based on the corresponding information extraction results to obtain the alignment matrix after registration processing includes:

[0015] Based on the corresponding information extraction results, ICP registration is performed between the first and third coloring images and between the second and third coloring images, respectively; and based on the registration results, one coloring image is selected from the first and second coloring images as the reference slice image; the other is the target slice image.

[0016] Based on the registration information of the reference slice image, the target slice image is transformed, thereby aligning the first slice, the second slice, and the H&E slice in the same coordinate system to obtain an alignment matrix.

[0017] Further, based on the corresponding information extraction results, ICP registration is performed between the first and third coloring images, and between the second and third coloring images, respectively; and based on the registration results, one coloring image is selected from the first and second coloring images as a reference slice image; the other is the target slice image, including:

[0018] Based on the cell nuclear coordinate data in the first and third staining images, two rounds of cyclic ICP registration calculations were performed on the first and third staining images, and the root mean square error value between the first and third staining images after the two rounds of ICP registration calculations was obtained; wherein, the first round of cyclic ICP registration calculations was performed in the original coordinate system, and the second round of cyclic ICP registration calculations was performed in the normalized coordinate system;

[0019] Similarly, based on the cell nuclear coordinate data in the second and third staining images, two rounds of cyclic ICP registration calculations were performed on the second and third staining images, and the root mean square error value between the second and third staining images after the two rounds of ICP registration calculations was obtained.

[0020] The reference slice and the target slice are determined based on the root mean square error between the first and third staining images and the root mean square error between the second and third staining images.

[0021] Furthermore, the stopping conditions for the two-round cyclic ICP registration calculation include: the number of cycles reaches a preset number or the root mean square error value is less than a preset root mean square error threshold, wherein the preset root mean square error threshold in the first round of cyclic ICP registration calculation is greater than the preset root mean square error threshold in the second round of cyclic ICP registration calculation.

[0022] Furthermore, the quantitative statistical analysis of the deviation coefficient of the positive rate of the same cancerous region in the first slice and the second slice based on the alignment matrix includes:

[0023] Based on the alignment matrix, calculate the deviation coefficient of the positive rate of the same cancerous region in the first slice and the second slice;

[0024] If the deviation coefficient of the positive rate is greater than the preset error value, the staining procedure of the first and second sections is considered unqualified.

[0025] If the difference in the positive rate is less than the preset error value, the staining procedure of the first and second sections is considered to be qualified.

[0026] Furthermore, the deviation coefficient of the positive rate of the same cancerous area in the first slice and the second slice is calculated using the following formula:

[0027] ;

[0028] in, The total number of selected grid cells; Let i be the number of cells in the i-th grid. The total number of cancer cells within the selected compliant grid; Let be the bias coefficient of a certain biomarker in the i-th grid.

[0029] Secondly, this application also provides a quantitative analysis system for cancerous regions of tissue sections, comprising:

[0030] The acquisition module is used to acquire the staining images corresponding to the first slice, the second slice, and the H&E slice. The first slice and the second slice are two consecutive tissue slices from the same cancerous tissue. The staining image of the first slice is denoted as the first staining image, the staining image of the second slice is denoted as the second staining image, and the staining image of the H&E slice is denoted as the third staining image.

[0031] The information extraction module is used to extract image information from the first staining image, the second staining image, and the third staining image to obtain the information extraction result corresponding to each staining image. The information extraction results of the first staining image and the second staining image include: data of cell nuclear coordinates and fluorescence intensity of biomarkers in the staining image. The information extraction result of the third staining image includes: data of cell nuclear coordinates, fluorescence intensity of biomarkers, and information of cancerous areas in the staining image.

[0032] The registration module is used to perform ICP registration processing on the first, second, and third stained images based on the information extraction results corresponding to each stained image, and to obtain the alignment matrix after registration processing.

[0033] The quantitative analysis module is used to perform quantitative statistical analysis on the deviation coefficient of the positive rate of the same cancerous region in the first slice and the second slice based on the alignment matrix.

[0034] Thirdly, this application also provides a computer electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the quantitative analysis method for cancerous areas of tissue slices as described in any of the above-mentioned methods.

[0035] Fourthly, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method for quantitative analysis of cancerous regions in tissue slices as described above.

[0036] The present application provides a method, system, device, and medium for quantitative analysis of cancerous areas in tissue sections, the advantages of which are:

[0037] 1. Achieve precise alignment at the single-cell level across slices, breaking through the bottleneck of traditional registration accuracy:

[0038] This application uses the nuclear coordinate point cloud as the core data for registration, replacing the traditional image-level pixel / feature point matching, thus avoiding the interference of staining differences and imaging noise on the registration benchmark from the source. At the same time, it adopts a two-stage ICP algorithm of "coarse registration of the original coordinate system + fine registration of the [0,1] normalized coordinate system", which not only solves the initial position offset problem, but also eliminates scale differences through coordinate normalization, realizing single-cell-level spatial alignment (deviation ≤0.3μm) between the first and second slices and the H&E slice, meeting the ideal requirement of "global alignment of tissue scale + consistency of single-cell scale", and overcoming the defects of local drift and inaccurate cell-level correspondence in traditional registration methods.

[0039] 2. Enables precise localization of homologous cancerous regions across slices, improving the accuracy of quantitative analysis:

[0040] This application uses H&E slides (the gold standard in pathology) to locate cancerous areas and precisely synchronizes them to the first and second slides through coordinate mapping. This ensures that the statistical objects are "homologous cancerous areas with completely overlapping spatial locations," solving the problems of large errors in manual selection of cancerous areas and inaccurate correspondence across slide regions in traditional methods. At the same time, the positive rate statistics are based on single-cell level data, avoiding quantitative distortion caused by regional bias, and making the analysis results more consistent with the actual pathological state of the tissue. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 This is a flowchart illustrating the quantitative analysis method for cancerous areas of tissue sections in the embodiments of this application;

[0043] Figure 2 These are comparison images of point cloud effects before and after registration in the embodiments of this application;

[0044] Figure 3 This is a diagram showing the mesh analysis results in an embodiment of this application;

[0045] Figure 4 This is a schematic diagram of the structure of a quantitative analysis system for cancerous areas of tissue sections according to an embodiment of this application;

[0046] Figure 5 This is a schematic diagram of the structure of a computer electronic device according to an embodiment of this application. Detailed Implementation

[0047] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0048] It should be noted that when an element is said to be "fixed" to another element, it can be directly on the other element or there may be an intervening element. When an element is said to be "connected" to another element, it can be directly connected to the other element or there may be an intervening element. Conversely, when an element is said to be "directly" on another element, there is no intervening element. The terms "vertical," "horizontal," "left," "right," and similar expressions used in this document are for illustrative purposes only.

[0049] In this application, unless otherwise expressly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0050] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0051] The terminology used in one or more embodiments of this application is for the purpose of describing particular embodiments only and is not intended to limit the scope of one or more embodiments of this application. The singular forms “a,” “the,” and “the” used in one or more embodiments of this application are also intended to include the plural forms unless the context clearly indicates otherwise.

[0052] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein in the template description is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0053] It should be understood that although the terms first, second, etc., may be used to describe various information in one or more embodiments of this application, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, first may also be referred to as second without departing from the scope of one or more embodiments of this application, and similarly, second may also be referred to as first. Depending on the context, the word "if" as used herein may be interpreted as "when" or "when".

[0054] Please see Figure 1 This application provides a method for quantitative analysis of cancerous regions in tissue sections, comprising at least the following steps:

[0055] S10. Obtain the staining images corresponding to the first slice, the second slice, and the H&E slice, wherein the first slice and the second slice are two consecutive tissue slices from the same cancerous tissue, the staining image of the first slice is denoted as the first staining image, the staining image of the second slice is denoted as the second staining image, and the staining image of the H&E slice is denoted as the third staining image.

[0056] Specifically, before conducting this experiment, it is necessary to prepare the first section, the second section, and the H&E section. It should be noted that the first section and the second section are adjacent and continuous sections (the thickness is usually 4-5 μm to ensure the continuity of the tissue structure and reduce spatial displacement across sections), and the H&E section is the corresponding section of the tissue block (it can be adjacent to the first / second section or separated by no more than 2 sections to ensure that the cancerous area is homologous).

[0057] The staining logic for the first slice, the second slice, and the H&E slice is as follows:

[0058] First, the first and second sections were stained using fluorescence staining to target biomarkers (such as PD-L1, Ki-67, and other markers related to tumor proliferation / immunity). DAPI nuclear staining could also be added (as a stable "nuclear localization marker," avoiding staining differences from affecting localization), ultimately generating the first and second staining images (fluorescence modality). H&E sections were stained using conventional hematoxylin-eosin staining to generate the third staining image (bright field modality). The advantage of H&E staining is its ability to clearly distinguish pathological structures such as tumor parenchyma, stroma, and necrotic areas, providing a "gold standard reference" for subsequent "precise selection of cancerous areas."

[0059] Finally, after staining, the three slides were scanned using a high-resolution digital pathology scanner (such as a whole-slide imaging system, WSI) to ensure that the imaging parameters were consistent (exposure time, resolution, and magnification were uniform to reduce batch effects), and stored in a lossless format such as TIFF to preserve the original data of cell nuclear details and fluorescence intensity.

[0060] S20. Image information is extracted from the first staining image, the second staining image, and the third staining image to obtain the information extraction result corresponding to each staining image. The information extraction results of the first staining image and the second staining image include: data of cell nuclear coordinates and fluorescence intensity of biomarkers in the staining image. The information extraction results of the third staining image include: data of cell nuclear coordinates, fluorescence intensity of biomarkers, and information of cancerous areas in the staining image.

[0061] In one embodiment of this application, the StrataQuest analysis software is used to extract image information from the first staining image, the second staining image, and the third staining image to obtain the corresponding information extraction results.

[0062] It should be noted that StrataQuest is a tissue flow cytometry quantitative analysis platform developed by TissueGnostics (Austria), specifically designed for multi-parameter quantitative analysis at the pathological section, tissue microenvironment, and single-cell level.

[0063] Specifically, in this embodiment, the StrataQuest analysis software is used to extract image information from the first staining image (first fluorescent section), the second staining image (second fluorescent section), and the third staining image (H&E section). During import, channel calibration is first performed (fluorescent sections distinguish between the DAPI reference channel and the biomarker signal channel, and H&E sections separate the hematoxylin channel) and a unified analysis template is created to avoid batch effects. Subsequently, through adaptive threshold segmentation, watershed algorithm, and morphological screening, the effective cell nuclei in the three staining images are accurately segmented and their physical spatial coordinates (μm units) are calculated. For fluorescent sections, an adaptation ROI is generated centered on the cell nucleus. After background subtraction, the average fluorescence intensity of the target biomarker is extracted and bound to the corresponding cell nucleus coordinates. Finally, quality control is completed through visualization verification and outlier data removal, and a standardized structured dataset containing cell nucleus coordinates, fluorescence intensity, and other information is output. This data can be directly adapted for subsequent ICP registration, providing support for precise alignment at the single-cell level.

[0064] Understandably, the third staining map also includes information about cancerous areas.

[0065] S30. Based on the information extraction results corresponding to each coloring image, perform ICP registration processing on the first coloring image, the second coloring image, and the third coloring image to obtain the alignment matrix after registration processing.

[0066] Specifically, the core approach employs a two-stage design of "coarse registration of the original coordinate system + fine registration of the normalized coordinate system" to eliminate the impact of scale differences on alignment accuracy. Combined with registration error screening to select stable benchmark slices, the coordinate system of the three slices is ultimately unified, ensuring single-cell-level spatial consistency. The specific process is as follows:

[0067] In one embodiment of this application, step S30 includes:

[0068] S301. Based on the corresponding information extraction results, perform ICP registration processing between the first and third coloring images and between the second and third coloring images respectively; and based on the registration processing results, select one coloring image from the first and second coloring images as the reference slice image; the other one is the target slice image.

[0069] Specifically, in this embodiment, the cell nucleus point cloud of the H&E slice is used as a fixed reference. Registration calculations are performed on the first slice and the second slice in two rounds. The reference slice is locked through comprehensive error evaluation, laying a reliable foundation for subsequent unified coordinates.

[0070] In one specific embodiment of this application, step S301 includes:

[0071] S3011. Based on the cell nuclear coordinate data in the first and third staining images, perform two rounds of cyclic ICP registration calculations on the first and third staining images, and obtain the root mean square error value between the first and third staining images after the two rounds of ICP registration calculations; wherein, the first round of cyclic ICP registration calculations is performed in the original coordinate system, and the second round of cyclic ICP registration calculations is performed in the normalized coordinate system.

[0072] Specifically, using the nuclear coordinate point cloud of the first staining image (P1, extracted from S20) and the nuclear coordinate point cloud of the third staining image (P3, H&E slice, as the baseline) as input, ICP registration is performed in two iterations. The specific operation is as follows:

[0073] First round of ICP registration:

[0074] The cell nucleus coordinate point clouds of the first, second, and H&E sections were obtained. All coordinates retained their original physical scale (in micrometers) to accurately reflect the spatial relationship of the tissue sections.

[0075] Coarse registration of the first slice with H&E:

[0076] 1.1 Using the nucleus point cloud of the H&E slice as a reference, and taking the coordinates of each nucleus in the first slice point cloud as the reference, find the nearest matching point in the H&E slice point cloud to form an initial set of matching pairs;

[0077] 1.2 Solve the rigid transformation matrix using the least squares method. This matrix contains translation and rotation angle parameters, which can minimize the overall distance between all points in the first slice and their corresponding matching points after transformation.

[0078] 1.3 Iterative optimization: Update the coordinates of the first slice according to the solved transformation matrix, find matching points again and calculate the new transformation matrix. Repeat this process until the root mean square error of two adjacent iterations does not exceed 0.2 micrometers. Stop the iteration and record the coarse registration error (root mean square error) at this time.

[0079] Second round of ICP registration:

[0080] By normalizing coordinates to unify the magnitude, we can more accurately capture minute positional deviations and improve alignment precision.

[0081] Coordinate normalization: The point cloud coordinates of the first slice, the second slice, and the H&E slice are normalized. Specifically, for the x and y coordinates of each point cloud, the minimum value of the corresponding coordinate is subtracted from the coordinate value, and then divided by the difference between the maximum and minimum values ​​of the coordinate. Finally, all coordinates are uniformly mapped to the range of 0 to 1. At the same time, the bidirectional mapping relationship between the original coordinates and the normalized coordinates is preserved to facilitate the subsequent restoration of the physical scale.

[0082] Precise registration of the first slice with the H&E slice:

[0083] 2.1. Using the point cloud in the normalized H&E slice as a reference, perform ICP calculation on the normalized point cloud of the first slice. Since the coordinate magnitude has been unified, it can more sensitively identify and correct small positional offsets.

[0084] 2.2 Similarly, solve the rigid transformation matrix, and set the iteration stopping condition to the error change in the normalized coordinate system not exceeding 0.001 (this value corresponds to about 0.1 micrometers in the original physical coordinates). After completion, convert the fine registration error back to the original physical scale and record it.

[0085] Fine registration of the second slice with the H&E slice: Following the same logic as the first slice, perform ICP registration on the normalized point cloud of the second slice and the normalized point cloud of the H&E slice, solve the transformation matrix and record the fine registration error after transformation.

[0086] S3012. Similarly, based on the cell nuclear coordinate data in the second and third staining images, two rounds of cyclic ICP registration calculations are performed on the second and third staining images, and the root mean square error value between the second and third staining images after the two rounds of ICP registration calculations is obtained.

[0087] Specifically, following the registration logic of the first slice and the H&E slice, the second slice and the H&E slice are registered in the same two rounds of ICP, the transformation matrix is ​​solved and the fine registration error after transformation is recorded.

[0088] It should be noted that when performing two rounds of cyclic ICP registration calculation, it is necessary to set the conditions for stopping the loop. In one embodiment of this application, the conditions for stopping the loop of two rounds of cyclic ICP registration calculation include: the number of loops reaches a preset number of values ​​or the root mean square error value is less than a preset root mean square error threshold. In this case, the preset root mean square error threshold in the first round of cyclic ICP registration calculation is greater than the preset root mean square error threshold in the second round of cyclic ICP registration calculation.

[0089] S3013. Based on the root mean square error between the first and third staining maps and the root mean square error between the second and third staining maps, determine the reference slice map and the target slice map.

[0090] Specifically, through two rounds of error calculation, the error from the second round is taken as the final error. Reference slice selection: Comparing the final registration errors of the two slices, the slice with the smaller error indicates that its alignment stability with the H&E slice is stronger, and it is set as the "reference slice"; the other slice with a relatively larger error is used as the "target slice".

[0091] S302. Based on the registration information of the reference slice image, the target slice image is transformed to align the first slice, the second slice, and the H&E slice in the same coordinate system, thereby obtaining an alignment matrix.

[0092] Specifically, the transformation rules of the reference slice are extracted: the complete transformation relationship determined by the reference slice after two rounds of registration is retrieved. This relationship records the precise mapping rules from the original coordinates of the reference slice to the coordinates of the H&E slice, ensuring the consistency of physical scale.

[0093] Target slice coordinate transformation:

[0094] 3.1 First, using the two-round registration transformation relationship between the target slice itself and the H&E slice, the original coordinates of the target slice are initially mapped to the H&E slice coordinate system;

[0095] 3.2. Based on the final registration error between the reference slice and the H&E slice, fine-tune the initial transformation coordinates of the target slice to ensure that the positional deviation of a single cell nucleus in the H&E coordinate system can be set according to the actual situation, for example, it can be set to no more than 0.3 micrometers.

[0096] Output alignment results: After the above processing, the coordinates of all cell nuclei in the first and second slices are completely consistent with the coordinate system of the H&E slices. The same cell corresponds to the same spatial position in the three slices, providing a solid spatial benchmark for the subsequent accurate localization and quantitative analysis of the "cancer area". Figure 2 This is a comparison image of the point cloud effect before and after registration.

[0097] S40. Based on the alignment matrix, perform quantitative statistical analysis on the deviation coefficient of the positive rate of the same cancerous region in the first slice and the second slice.

[0098] Specifically, the quantitative analysis process in this step is as follows:

[0099] First, precisely select areas with the same cancerous lesions:

[0100] Based on the registered third staining map (H&E section), the target cancerous area (such as the tumor parenchyma area) is manually or automatically segmented (based on the cell density and morphological characteristics of H&E staining to distinguish between normal tissue or necrotic areas).

[0101] By using the registered coordinate mapping relationship, the cancerous region is synchronized to the registered first staining map (P1') and second staining map (P2'), ensuring that the two fluorescent sections represent "completely identical spatial regions".

[0102] Secondly, the criteria for determining a positive result can be established as follows:

[0103] Based on the fluorescence intensity data of biomarkers from the first and second staining images, a "negative control calibration method" was used to set a positive threshold: the mean fluorescence intensity of known negative regions (such as cell nuclei in normal tissue) in the slide plus 2 standard deviations was used as the threshold. Cell nuclei with values ​​above this threshold were judged as "positive cells", and those below were judged as "negative cells". This standard ensures the objectivity of the positive determination and avoids human experience errors.

[0104] Finally, quantitative statistics and analysis:

[0105] Basic statistics: The positive rate (number of positive cells / total number of cell nuclei in the region) of the same cancerous area in two fluorescent sections was calculated to reflect the difference in the expression level of biomarkers.

[0106] Consistency verification: Compare the positive rate results of the first and second sections, calculate the correlation between the two (such as the Sperman correlation coefficient), and verify the stability of biomarker expression in serial sections;

[0107] Results output: Generates a quantitative analysis report, including the spatial extent of the cancerous area, the positive rate values ​​and correlations of the two sections, and a heatmap of the distribution of positive cells, ensuring that the results are reproducible and comparable.

[0108] In one embodiment of this application, step S40 includes:

[0109] S401. Based on the alignment matrix, calculate the deviation coefficient of the positive rate of the same cancerous region in the first slice and the second slice:

[0110] If the deviation coefficient of the positive rate is greater than the preset error value, the staining procedure of the first and second sections is considered unqualified.

[0111] If the difference in the positive rate is less than the preset error value, the staining procedure of the first and second sections is considered to be qualified.

[0112] Specifically, this step serves as a "quality control checkpoint" in the entire quantitative analysis process. By comparing the differences in the positive rates of homologous cancerous regions in consecutive sections, it assesses the stability and accuracy of the staining process, preventing subsequent analytical conclusions from being distorted due to staining errors. The specific processing logic is as follows:

[0113] Based on the positive rate of "identical cancerous areas after registration", the numerical difference between the first and second sections was calculated to provide a quantitative basis for judging staining quality.

[0114] Calculation premise: The basis for difference calculation is the "same cancerous area" that is precisely matched after S30 registration - this area is a tissue area that is completely spatially overlapping in the first slice (reference / target slice) and the second slice (target / reference slice) (such as a sub-region of the tumor parenchyma, with an area of ​​≥0.1mm², containing at least 50 cell nuclei to ensure statistical reliability), and the positive rate statistical standards are completely consistent.

[0115] For the quantitative analysis phase, please refer to Figure 3 First, the cancerous area is manually or automatically divided into multiple grids, and these grids are then screened. We first randomly select grids that meet the condition (the number of cells in the grid is greater than a set threshold). Next, we calculate the coefficient of variation (CV) value for each biomarker in each selected grid. To synthesize the contributions of different grids, we use the proportion of cells in each grid relative to the total number of cells in all selected grids as a weight. This weight is multiplied by the CV value of each biomarker in the corresponding grid, and then summed across all grids to obtain the overall CV value of the tissue section. The specific calculation formula is as follows:

[0116] ;

[0117] in, The total number of selected grid cells; Let i be the number of cells in the i-th grid. The total number of cancer cells within the selected compliant grid; Let be the bias coefficient of a certain biomarker in the i-th grid.

[0118] It should be noted that before conducting quantitative statistical analysis of cancer grids, the cancer grids need to be screened first. For example, grids with cell counts below a cell count threshold or grids with a positive rate below a positive rate threshold need to be removed.

[0119] In this application, the cell number threshold and the positivity rate threshold can be set according to the actual situation. The settable parameters are as follows: grid size (in pixels), number of random sampling grids (randomly select grids from the slices for statistical analysis), positivity rate threshold (the positivity rate of biomarkers in the grid must be higher than this threshold to be considered a valid sample), CV threshold (bias coefficient threshold, default 0.3, if lower than this value, the results of the two slices are considered consistent), random seed (used to ensure the reproducibility of experimental results), and cell number threshold (the number of cells in the grid must be higher than this threshold to be considered a "valid sample" to avoid edge effects).

[0120] Judgment Logic and Result Interpretation:

[0121] If the deviation in the positive rate exceeds the preset error value, it indicates a significant deviation in the staining process between the first and second sections. This suggests a problem with the staining process, which may be due to: antibody incubation time being too long / too short for a particular section (leading to excessively strong / weak fluorescence), insufficient washing (non-specific staining interference), or abnormal exposure of the fluorescence channel. In such cases, the staining results cannot reflect the true expression of the biomarker and should be deemed "staining unqualified." It is recommended to re-prepare the sections and repeat the staining-analysis process.

[0122] If the difference in positive rates is less than or equal to the preset error value, it means that the staining difference between the two sections is within a reasonable technical error range. The staining process can be considered stable, and the results can truly reflect the expression status of biomarkers in the tissue. It is judged as "staining qualified". Its quantitative analysis results can be used for subsequent pathological mechanism research (such as spatial heterogeneity analysis of tumor microenvironment).

[0123] It should be noted that, in the final step, the method in this application will randomly select cells with a number of cancer cells exceeding a threshold, count the number of cancer cells in the selected cells, calculate the CV value and total CV value in the selected cells, and record all parameter settings during the experiment.

[0124] The method for quantitative analysis of cancerous areas in tissue sections provided in this application has the following advantages:

[0125] 1. Achieve precise alignment at the single-cell level across slices, breaking through the bottleneck of traditional registration accuracy:

[0126] This application uses the nuclear coordinate point cloud as the core data for registration, replacing the traditional image-level pixel / feature point matching, thus avoiding the interference of staining differences and imaging noise on the registration benchmark from the source. At the same time, it adopts a two-stage ICP algorithm of "coarse registration of the original coordinate system + fine registration of the [0,1] normalized coordinate system", which not only solves the initial position offset problem, but also eliminates scale differences through coordinate normalization, realizing single-cell-level spatial alignment (deviation ≤0.3μm) between the first and second slices and the H&E slice, meeting the ideal requirement of "global alignment of tissue scale + consistency of single-cell scale", and overcoming the defects of local drift and inaccurate cell-level correspondence in traditional registration methods.

[0127] 2. Enables precise localization of homologous cancerous regions across slices, improving the accuracy of quantitative analysis:

[0128] This application uses H&E slices (the gold standard in pathology) to locate cancerous areas and precisely synchronizes them to the first and second slices through coordinate mapping. This ensures that the statistical objects are "homologous cancerous areas with completely overlapping spatial locations," solving the problems of large errors in manual selection of cancerous areas and inaccurate correspondence across slice areas in traditional methods. At the same time, the positive rate statistics are based on single-cell level data, avoiding quantitative distortion caused by regional bias, and making the analysis results more consistent with the actual pathological state of the tissue.

[0129] Please see Figure 4 This application also provides a quantitative analysis system 200 for cancerous regions of tissue sections, comprising:

[0130] The acquisition module 201 is used to acquire the staining images corresponding to the first slice, the second slice, and the H&E slice, wherein the first slice and the second slice are two consecutive tissue slices from the same cancerous tissue, the staining image of the first slice is denoted as the first staining image, the staining image of the second slice is denoted as the second staining image, and the staining image of the H&E slice is denoted as the third staining image.

[0131] Information extraction module 202 is used to extract image information from the first staining image, the second staining image, and the third staining image to obtain the information extraction result corresponding to each staining image. The information extraction results of the first staining image and the second staining image include: data of cell nuclear coordinates and fluorescence intensity of biomarkers in the staining image. The information extraction results of the third staining image include: data of cell nuclear coordinates, fluorescence intensity of biomarkers, and information of cancerous areas in the staining image.

[0132] Registration module 203 is used to perform ICP registration processing on the first staining image, the second staining image, and the third staining image according to the information extraction results corresponding to each staining image, and obtain the alignment matrix after registration processing.

[0133] The quantitative analysis module 204 is used to perform quantitative statistical analysis on the deviation coefficient of the positive rate of the same cancerous region in the first slice and the second slice according to the alignment matrix.

[0134] Please see Figure 5 This application also provides a computer electronic device 300, including a memory 303 and a processor 302. The memory 303 stores a computer program, and when the processor executes the computer program, it implements the steps of the quantitative analysis method for cancerous areas of tissue slices described in any of the above claims.

[0135] Specifically, the electronic device 300 includes: a transceiver 301, a bus interface, and a processor 302. The processor 302 is used to acquire staining images corresponding to a first slice, a second slice, and a H&E slice, wherein the first slice and the second slice are two consecutive tissue slices from the same cancerous tissue; the staining image of the first slice is denoted as the first staining image, the staining image of the second slice is denoted as the second staining image, and the staining image of the H&E slice is denoted as the third staining image; image information is extracted from the first staining image, the second staining image, and the third staining image to obtain the information extraction result corresponding to each staining image. The information extraction results of the first and second staining images include: data on cell nuclear coordinates and fluorescence intensity of biomarkers in the staining images; the information extraction results of the third staining image include: data on cell nuclear coordinates, fluorescence intensity of biomarkers, and information on cancerous areas in the staining image; based on the information extraction results corresponding to each staining image, ICP registration processing is performed on the first, second, and third staining images to obtain a registration alignment matrix; based on the alignment matrix, quantitative statistical analysis is performed on the deviation coefficient of the positive rate of the same cancerous areas in the first and second slices.

[0136] In this embodiment of the application, the electronic device 300 further includes a memory 303. Figure 5 In this context, the bus architecture can include any number of interconnected buses and bridges, specifically linking various circuits together, represented by one or more processors (processor 302) and memory (memory 303). The bus architecture can also link various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. The bus interface provides an interface. The transceiver 301 can be multiple elements, including transmitters and receivers, providing a unit for communicating with various other devices over a transmission medium. The processor 302 is responsible for managing the bus architecture and general processing, and the memory 303 can store data used by the processor 302 during operation.

[0137] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method for quantitative analysis of cancerous areas in tissue slices described above.

[0138] In this embodiment, the computer-readable storage medium can be a non-volatile storage medium or a volatile storage medium. For example, the computer storage medium may include, but is not limited to, various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0139] In all examples shown and described herein, any specific values ​​should be interpreted as merely exemplary and not as limitations; therefore, other examples of exemplary embodiments may have different values.

[0140] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0141] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can also be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the flowcharts and block diagrams in the accompanying drawings show the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that, in alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0142] In addition, the functional modules or units in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0143] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, 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 a terminal device (which may be a smartphone, personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application.

[0144] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for quantitative analysis of cancerous regions in tissue sections, characterized in that, include: Obtain the staining images corresponding to the first slice, the second slice, and the H&E slice, wherein the first slice and the second slice are two consecutive tissue slices from the same cancerous tissue, the staining image of the first slice is denoted as the first staining image, the staining image of the second slice is denoted as the second staining image, and the staining image of the H&E slice is denoted as the third staining image. Image information is extracted from the first staining image, the second staining image, and the third staining image to obtain the information extraction results corresponding to each staining image. The information extraction results of the first staining image and the second staining image include: data of cell nuclear coordinates and fluorescence intensity of biomarkers in the staining image. The information extraction results of the third staining image include: data of cell nuclear coordinates, fluorescence intensity of biomarkers, and information of cancerous areas in the staining image. Based on the information extraction results corresponding to each coloring image, ICP registration processing is performed on the first coloring image, the second coloring image, and the third coloring image to obtain the alignment matrix after registration processing. Based on the alignment matrix, a quantitative statistical analysis is performed on the deviation coefficient of the positive rate of the same cancerous region in the first slice and the second slice.

2. The method for quantitative analysis of cancerous areas in tissue sections according to claim 1, characterized in that, The step of extracting image information from the first, second, and third stained images to obtain corresponding information extraction results includes: The first, second, and third stained images were analyzed using StrataQuest software to extract image information and obtain the corresponding information extraction results.

3. The method for quantitative analysis of cancerous areas in tissue sections according to claim 1, characterized in that, The step of performing ICP registration on the first, second, and third coloring images based on the corresponding information extraction results to obtain the alignment matrix after registration includes: Based on the corresponding information extraction results, ICP registration is performed between the first and third coloring images and between the second and third coloring images, respectively; and based on the registration results, one coloring image is selected from the first and second coloring images as the reference slice image; the other is the target slice image. Based on the registration information of the reference slice image, the target slice image is transformed, thereby aligning the first slice, the second slice, and the H&E slice in the same coordinate system to obtain an alignment matrix.

4. The method for quantitative analysis of cancerous areas in tissue sections according to claim 3, characterized in that, Based on the corresponding information extraction results, ICP registration is performed between the first staining image and the third staining image, and between the second staining image and the third staining image, respectively. Based on the registration process, one staining image is selected from the first and second staining images as a reference slice image. The other image is a slice of the target image, including: Based on the cell nuclear coordinate data in the first and third staining images, two rounds of cyclic ICP registration calculations were performed on the first and third staining images, and the root mean square error value between the first and third staining images after the two rounds of ICP registration calculations was obtained; wherein, the first round of cyclic ICP registration calculations was performed in the original coordinate system, and the second round of cyclic ICP registration calculations was performed in the normalized coordinate system; Similarly, based on the cell nuclear coordinate data in the second and third staining images, two rounds of cyclic ICP registration calculations were performed on the second and third staining images, and the root mean square error value between the second and third staining images after the two rounds of ICP registration calculations was obtained. The reference slice and the target slice are determined based on the root mean square error between the first and third staining images and the root mean square error between the second and third staining images.

5. The method for quantitative analysis of cancerous areas in tissue sections according to claim 3, characterized in that, The stopping conditions for the two-round cyclic ICP registration calculation include: the number of cycles reaches a preset number or the root mean square error value is less than a preset root mean square error threshold. In the first round of cyclic ICP registration calculation, the preset root mean square error threshold is greater than the preset root mean square error threshold in the second round of cyclic ICP registration calculation.

6. The method for quantitative analysis of cancerous areas in tissue sections according to claim 1, characterized in that, The quantitative statistical analysis of the deviation coefficient of the positive rate of the same cancerous region in the first slice and the second slice according to the alignment matrix includes: Based on the alignment matrix, calculate the deviation coefficient of the positive rate of the same cancerous region in the first slice and the second slice; If the deviation coefficient of the positive rate is greater than the preset error value, the staining procedure of the first and second sections is considered unqualified. If the deviation coefficient of the positive rate is less than the preset error value, then the staining procedure of the first and second sections is considered qualified.

7. The method for quantitative analysis of cancerous areas in tissue sections according to claim 1, characterized in that, The deviation coefficient of the positive rate of the same cancerous area in the first slice and the second slice is calculated using the following formula: ； in, The total number of selected grid cells; Let i be the number of cells in the i-th grid. The total number of cancer cells within the selected compliant grid; Let be the bias coefficient of a certain biomarker in the i-th grid.

8. A quantitative analysis system for cancerous regions in tissue sections, characterized in that, include: The acquisition module is used to acquire the staining images corresponding to the first slice, the second slice, and the H&E slice. The first slice and the second slice are two consecutive tissue slices from the same cancerous tissue. The staining image of the first slice is denoted as the first staining image, the staining image of the second slice is denoted as the second staining image, and the staining image of the H&E slice is denoted as the third staining image. The information extraction module is used to extract image information from the first staining image, the second staining image, and the third staining image to obtain the information extraction result corresponding to each staining image. The information extraction results of the first staining image and the second staining image include: data of cell nuclear coordinates and fluorescence intensity of biomarkers in the staining image. The information extraction result of the third staining image includes: data of cell nuclear coordinates, fluorescence intensity of biomarkers, and information of cancerous areas in the staining image. The registration module is used to perform ICP registration processing on the first, second, and third stained images based on the information extraction results corresponding to each stained image, and to obtain the alignment matrix after registration processing. The quantitative analysis module is used to perform quantitative statistical analysis on the deviation coefficient of the positive rate of the same cancerous region in the first slice and the second slice based on the alignment matrix.

9. A computer electronic device, characterized in that, The method includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps of the method for quantitative analysis of cancerous areas in tissue sections according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the method for quantitative analysis of cancerous regions in tissue sections according to any one of claims 1-7.

Images