Method for dynamic evaluation of endothelial barrier integrity in an organ-on-chip

Abstract

The present application relates to the field of image processing, in particular to a kind of dynamic evaluation method and system of endothelial barrier integrity in organ chip.The method obtains the dual-channel fluorescence image of cell nucleus and intercellular junction protein by confocal microscope, and extracts four key feature parameters based on image processing technology: cell coverage, cell nucleus distribution uniformity, intercellular junction protein intensity and junction protein integrity.Through the fusion of these parameters, a comprehensive evaluation index is obtained, and compared with the preset threshold value, to determine whether the endothelial barrier is complete.The present application avoids the interference of traditional permeability detection method, realizes the multidimensional, quantitative and high reliability evaluation of barrier structure, and is suitable for various organ chip models.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of image processing, in particular to a method for dynamically evaluating endothelial barrier integrity in an organ chip. BACKGROUND

[0002] The rapid development of organ chip technology puts forward higher requirements for the functional evaluation of in vitro biological models. As a core functional unit, the integrity of the endothelial barrier is crucial. One of the current standard methods in this field is based on the detection of fluorescent dye penetration function. This method requires the addition of exogenous tracer dye to the chip chamber during detection, and the leakage amount is measured in real time to indirectly evaluate the barrier function. However, this method has inherent defects: first, the addition of exogenous dye itself pollutes the culture system, and it may interfere with cell activity or be toxic during penetration, affecting the functional state of the barrier itself, resulting in inaccurate leakage data that cannot accurately reflect the true physiological condition of the barrier; second, this method relies on a single penetration indicator, and the judgment standard is rigid, making it difficult to adapt to different types and different culture standards of organ chip models.

[0003] The present application proposes an evaluation strategy based on direct morphological analysis. This strategy focuses on the direct observation and quantitative analysis of cell structure stability and coverage by specifically fluorescent staining endothelial barrier cells, shifting the evaluation focus from dynamic function measurement that may be disturbed to direct observation and quantitative analysis of cell structure stability and coverage. The advantage of this method is that even if the staining process may cause cell activity to stop, it maximally preserves the complete morphological structure of the sample at the instant of staining. This structural information is stable and not affected by subsequent analysis under visual conditions, providing intuitive and reliable data basis for evaluation. More importantly, this method can comprehensively reflect the structural state of the barrier by extracting multiple parameters such as cell coverage, distribution uniformity, and connection protein strength and integrity, and can flexibly adjust the weight and judgment threshold of each parameter according to the specific requirements of different types of organ chips, significantly enhancing the universality and adaptability. SUMMARY

[0004] The present application provides a method for dynamically evaluating endothelial barrier integrity in an organ chip to solve the existing problems: the existing organ chip endothelial barrier evaluation technology based on dye penetration method is inaccurate, incomplete and has poor adaptability due to exogenous dye interference and reliance on a single functional indicator.

[0005] The organ chip endothelial barrier integrity dynamic evaluation method of the present application adopts the following technical scheme:

[0006] In the first aspect of the present application, a method for dynamically evaluating endothelial barrier integrity in an organ chip is provided, which comprises the following steps:

[0007] Acquire high-resolution images of endothelial barrier cells under fluorescence channel and transmission light channel respectively using a confocal microscope, and label the intercellular junction protein and the cell nucleus for double staining and mark the pixel points in the cell nucleus region and the pixel points in the intercellular junction protein region, and obtain a fluorescence image simultaneously marking the pixel points in the cell nucleus region and the pixel points in the intercellular junction protein region;

[0008] Cluster the pixel points in the endothelial barrier cell transmission light image into a high gray value class and a low gray value class according to the gray value of the endothelial barrier cell transmission light image, and obtain the cell coverage W according to the ratio of the number of pixel points in the low gray value class to the number of all pixel points in the endothelial barrier cell transmission light image;

[0009] Detect the position and the number of pixel points of the cell nucleus region in the fluorescence image, obtain the cell nucleus regions having an adjacent relationship with each cell nucleus region by using a Delaunay triangular net, take the farthest distance of all adjacent cell nucleus regions as the side length of the window, divide the image into N windows, and obtain the cell nucleus distribution uniformity parameter according to the information entropy of the number of cell nucleus regions in different windows;

[0010] Select the non-cell nucleus region between the nearest cell nucleus region connected domain having an adjacent relationship with each cell nucleus region, and record the region as the intercellular junction protein template region between two adjacent cell nuclei, take the mean value of the gray values of all pixel points marked as intercellular junction protein corresponding pixel points in the intercellular junction protein template region as the intercellular junction protein intensity of the two adjacent cell nuclei, calculate the Pearson correlation coefficient of the intercellular distance and the intercellular junction protein intensity and normalize it, and obtain the intercellular junction protein intensity parameter;

[0011] Set an image grid, screen all grids having pixel points marked as intercellular junction protein, take the pixel point with the lowest gray value of the pixel points marked as intercellular junction protein in each grid as a seed point, take the gray value of the eight-neighborhood pixel points of the seed point as the growth condition and update the position of the seed point, find the suspected weak region connected domain in each grid, and obtain the intercellular junction protein integrity parameter according to the ratio of the number of pixel points in the suspected weak region connected domain in each grid to the number of pixel points in the grid;

[0012] Obtain the endothelial barrier integrity probability parameter by combining the cell coverage, the cell nucleus distribution uniformity parameter, the intercellular junction protein intensity parameter and the intercellular junction protein integrity parameter;

[0013] Set the endothelial barrier integrity probability parameter threshold, compare the endothelial barrier integrity probability parameter with the endothelial barrier integrity probability parameter threshold, and judge whether the endothelial barrier is complete.

[0014] Further, the confocal microscope is used to collect high-resolution images of the endothelial barrier cells under fluorescence channel and transmission light channel respectively, and intercellular junction protein and cell nucleus are double-labeled and marked as pixel points of the cell nucleus region and pixel points of the intercellular junction protein region, and a fluorescence image of the pixel points of the cell nucleus and the intercellular junction protein region is obtained, and the specific method comprises:

[0015] The endothelial barrier cells are double-labeled by using an antibody for a specific intercellular junction protein and DAPI nuclear dye, and then high-resolution images are collected under fluorescence channel and transmission light channel respectively by using a confocal microscope, and all shooting parameters are ensured to be consistent. The size of the image is 400*400 pixels, and the region with fluorescence reaction in the gray scale image scanned by 405nm laser is marked as the cell nucleus region, and the region with fluorescence reaction in the gray scale image scanned by 488nm laser is marked as the intercellular protein region, and a transmission light image of the endothelial barrier cells and a fluorescence image of the endothelial barrier cells are obtained.

[0016] Further, the pixel points of the transmission light image of the endothelial barrier cells are clustered into a high gray value class and a low gray value class according to the gray value of the transmission light image of the endothelial barrier cells, and the cell coverage W is obtained according to the ratio of the number of pixel points in the low gray value class to the number of all pixel points in the transmission light image of the endothelial barrier cells, and the specific method comprises:

[0017] The gray values of the pixel points in the transmission light image of the endothelial barrier cells are counted, and the maximum value and the minimum value thereof are taken as initial clustering centers, each pixel point is divided into a class corresponding to the clustering center with which the gray value is closer, the gray mean value of all pixel points in each class is recalculated, and the new clustering center is updated, and the pixel division and clustering center updating process is repeated until the iteration termination condition is met: after a certain iteration, the change amount of the gray values of the two newly calculated clustering centers compared with the previous iteration is less than 1% of the gray value of the current clustering center; based on the last pixel division result, all pixel points in the image are clustered into two classes of high gray value and low gray value, and the ratio of the number of pixel points in the low gray value class to the number of pixel points in the entire transmission light image of the endothelial barrier cells is taken as the cell coverage W, and if the ratio of the number of pixel points in the class with fewer pixel points to the number of pixel points in the entire transmission light image of the endothelial barrier cells is less than or equal to 5%, the cell coverage W=1.

[0018] Further, the position and the number of pixel points of the cell nucleus region in the fluorescence image are detected, the cell nucleus regions having an adjacent relationship with each other are obtained by using a Delaunay triangular net, the farthest distance of all adjacent cell nucleus regions is taken as the side length of a window, the image is divided into N windows, and the cell nucleus distribution uniformity parameter is obtained according to the information entropy of the number of cell nucleus regions in different windows, and the specific method comprises: ​

[0019] In the fluorescence image, the geometric center of each pixel point connected domain labeled as a nucleus region is calculated as the center point of each nucleus. A triangular mesh is constructed based on the coordinates of all nucleus center points by applying a Delaunay triangulation algorithm. The adjacent relationship between the nucleus center pairs is determined according to the connection relationship formed by the triangulation, and all pixel point connected domains labeled as nucleus regions having an adjacent relationship are obtained. The maximum distance in all nucleus center pairs is calculated, and the maximum distance is taken as the cell distance parameter. As the side length of the analysis window, the entire fluorescence image is segmented into windows of The number of nucleus region pixel point connected domains in each window is recorded. The uniformity degree parameter of the nucleus distribution is obtained by the normalized entropy of the number distribution of the nucleus region pixel point connected domains in different windows. The specific method is as follows:

[0020]

[0021] In the formula, N represents the number of windows, N represents the total number of nucleus region pixel point connected domains, N represents the number of nucleus region pixel point connected domains in the th window.

[0022] Further, the non-nucleus region between the nearest nucleus region connected domain having an adjacent relationship with each nucleus region is selected and recorded as the connecting protein template region between the two adjacent nuclei. The average gray value of all pixel points labeled as connecting protein corresponding pixel points in the connecting protein template region is taken as the intercellular connecting protein intensity of the two adjacent nuclei. The Pearson correlation coefficient between the intercellular distance and the intercellular connecting protein intensity is calculated and normalized to obtain the intercellular connecting protein intensity parameter. The specific method includes:

[0023] The geometric center of the nearest nucleus region connected domain having an adjacent relationship with each nucleus region is connected as a reference line. A series of parallel lines are generated by pixel-by-pixel translation in the vertical direction to both sides of the reference line until the newly generated parallel line no longer intersects with two nucleus region connected domains at the same time. The first intersection point passing through two nuclei and two nucleus region connected domains at the same time is recorded. The line segment between the two points is recorded as the effective connection region on the parallel line. All pixel points on all effective connection regions are collected as the connecting protein template region between the two adjacent nuclei. The average gray value of all pixel points labeled as connecting protein corresponding pixel points in the connecting protein template region is taken as the intercellular connecting protein intensity of the two adjacent nuclei. The Pearson correlation coefficient between the intercellular distance and the intercellular connecting protein intensity is calculated and normalized to obtain the intercellular connecting protein intensity parameter. The specific method is as follows:

[0024] ;

[0025] ;

[0026] wherein, represents the linear relationship between the nuclear distance and the intensity of the connexin protein, represents the total number of the pixel point connected domains in the nuclear region, represents the distance between the geometric center of the i-th pixel point connected domain in the nuclear region and the most adjacent pixel point connected domain in the nuclear region, represents the intensity of the connexin protein between the i-th pixel point connected domain in the nuclear region and the most adjacent pixel point connected domain in the nuclear region, represents the average of the distance between the geometric center of each pixel point connected domain in the nuclear region and the most adjacent pixel point connected domain in the nuclear region, represents the average of the intensity of the connexin protein between each pixel point connected domain in the nuclear region and the most adjacent pixel point connected domain in the nuclear region, represents the intensity parameter of the connexin protein.

[0027] Further, the image grid is set, and the grid in which the pixel point marked as the connexin protein exists is screened in all the grids. The pixel point with the lowest gray value marked as the connexin protein pixel point in each grid is taken as a seed point. The gray value of the pixel point in the eight-neighborhood of the seed point is taken as a growth condition, and the position of the seed point is updated. The connected domain of the suspected weak region in each grid is found. The ratio of the number of the pixel points in the connected domain of the suspected weak region in each grid to the number of the pixel points in the grid is taken as the integrity parameter of the connexin protein. The specific method includes:

[0028] The image grid is set in the fluorescence image. The fluorescence image is evenly divided into grids with a size of The grid in which the pixel point of the connexin protein is covered is recorded as the connexin protein grid. The average of the gray values of all the connexin protein pixel points in the y-th grid is calculated The connexin protein pixel point with the lowest gray value in each connexin protein grid is recorded as the first pixel point of the suspected weak region. If there is a pixel point with a gray value lower than in the eight-neighborhood of the first pixel point, the pixel point with a gray value lower than is marked as the pixel point of the suspected weak region in the grid. It is judged whether there is a pixel point with a gray value lower than in the eight-neighborhood of the pixel point newly marked as the pixel point of the suspected weak region. If there is a pixel point with a gray value lower than , the pixel point with a gray value lower than The pixel points in the grid are marked as the pixel points of the suspected weak area in the grid. The above eight-neighborhood judgment and suspected weak area pixel point marking process is repeated until all pixel points in the grid are judged. The intercellular connection protein integrity parameter is calculated according to the following method:

[0029]

[0030] The intercellular connection protein integrity parameter is represented by P. The total number of grids covering the intercellular connection protein pixel points is represented by N. The number of suspected weak area pixel points in the i-th grid is represented by Ni. The number of intercellular connection protein pixel points in the i-th grid is represented by Ni.

[0031] Further, the endothelial barrier integrity probability parameter is obtained by combining the cell coverage, the cell nucleus distribution uniformity parameter, the intercellular connection protein strength parameter, and the intercellular connection protein integrity parameter, and the specific method comprises:

[0032]

[0033] P represents the endothelial barrier integrity probability parameter, W represents the cell coverage, represents the cell nucleus distribution uniformity parameter, represents the intercellular connection protein strength parameter, represents the intercellular connection protein integrity parameter.

[0034] Further, the threshold of the endothelial barrier integrity probability parameter is set, and the endothelial barrier is judged to be complete by comparing the endothelial barrier integrity probability parameter with the threshold of the endothelial barrier integrity probability parameter, and the specific method comprises:

[0035] According to the specific organ chip application, the corresponding threshold is set As a specific requirement for evaluating the endothelial barrier integrity, if the endothelial barrier integrity probability parameter is greater than the set threshold , it is considered that the endothelial barrier integrity is good, and if the endothelial barrier integrity probability parameter is less than or equal to the set threshold , it is considered that the endothelial barrier integrity is poor.

[0036] ​​​​In a second aspect, the application provides a system for dynamically evaluating endothelial barrier integrity in an organ-on-a-chip, comprising a data acquisition module, a cell coverage calculation module, a cell nucleus distribution uniformity parameter calculation module, a connexin intensity parameter calculation module, a connexin integrity parameter calculation module, an endothelial barrier integrity probability parameter calculation module, and an endothelial barrier integrity judgment module, wherein:

[0037] The data acquisition module is configured to use a confocal microscope to acquire high-resolution images of endothelial barrier cells under a fluorescence channel and a transmission light channel respectively, to perform double-labeling staining on the connexin and the cell nucleus, to mark the cell nucleus region pixel points and the connexin region pixel points, and to obtain fluorescence images of the simultaneously labeled cell nucleus and connexin region pixel points.

[0038] The cell coverage calculation module is configured to cluster the pixel points of the endothelial barrier cell transmission light image into a high gray value class and a low gray value class according to the gray value of the endothelial barrier cell transmission light image, and to obtain the cell coverage W according to the ratio of the number of pixel points in the low gray value class to the number of all pixel points in the endothelial barrier cell transmission light image.

[0039] The cell nucleus distribution uniformity parameter calculation module is configured to detect the position and the number of pixel points of the cell nucleus region in the fluorescence image, to obtain cell nucleus regions having an adjacent relationship with each cell nucleus region using a Delaunay triangular net, to take the farthest distance of all adjacent cell nucleus regions as the window side length, to divide the image into N windows, and to obtain the cell nucleus distribution uniformity parameter according to the information entropy of the number of cell nucleus regions in different windows.

[0040] The connexin intensity parameter calculation module is configured to select a non-cell nucleus region between the nearest cell nucleus region connected domain having an adjacent relationship with each cell nucleus region, to record the region as a connexin template region between two adjacent cell nuclei, to take the mean value of the gray values of all pixel points marked as connexin corresponding pixel points in the connexin template region as the intercellular connexin intensity of the two adjacent cell nuclei, to calculate the Pearson correlation coefficient of the intercellular distance and the intercellular connexin intensity and normalize it, and to obtain the intercellular connexin intensity parameter.

[0041] The connexin integrity parameter calculation module is configured to screen all grids in which there are pixel points marked as intercellular connexin, to take the pixel point with the lowest gray value of the pixel points marked as intercellular connexin in each grid as a seed point, to update the seed point position according to the gray value size of the eight-neighborhood pixel points of the seed point as the growth condition, to find a suspected weak region connected domain in each grid, and to obtain the connexin integrity parameter according to the ratio of the number of pixel points in the suspected weak region connected domain in each grid to the number of pixel points in the grid.

[0042] An endothelial barrier integrity probability parameter calculation module is configured to obtain an endothelial barrier integrity probability parameter based on the cell coverage, the nucleus distribution uniformity parameter, the intercellular junction protein intensity parameter, and the intercellular junction protein integrity parameter.

[0043] An endothelial barrier integrity judgment module is configured to set an endothelial barrier integrity probability parameter threshold, and compare the endothelial barrier integrity probability parameter with the endothelial barrier integrity probability parameter threshold to determine whether the endothelial barrier is complete.

[0044] In a third aspect, the application provides a computer readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the steps of the organ-on-chip endothelial barrier integrity dynamic evaluation method.

[0045] In a fourth aspect, the application provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the organ-on-chip endothelial barrier integrity dynamic evaluation method when executing the computer program.

[0046] The technical scheme of the application has the following beneficial effects:

[0047] The automatic recognition and quantification of the cell coverage area are achieved, the subjective errors of manual interpretation are avoided, and the objectivity and repeatability of the evaluation are improved.

[0048] The precise distinction between the nucleus and the junction protein is ensured through the dual-channel fluorescence imaging and the specific wavelength marking, and reliable data basis is provided for subsequent analysis.

[0049] The adaptive clustering method is adopted to distinguish the cells from the background, the noise interference is effectively excluded, and the accuracy of the cell coverage calculation is improved.

[0050] The Delaunay triangulation and the information entropy are combined to objectively quantify the nucleus distribution uniformity, the subjectivity of manual division is avoided, and the reliability of the analysis is enhanced.

[0051] The distance-intensity correlation is calculated by constructing the junction protein template area, the functional compensation ability of the intercellular junction is revealed, and a dynamic functional evaluation index is provided.

[0052] The gridding and region growing algorithm are adopted to actively identify the local weak area, the shortcomings of the overall average analysis are made up, and the sensitivity of the defect detection is improved.

[0053] The barrier state is comprehensively evaluated through the multi-parameter fusion, the limitations of a single index are avoided, and the comprehensiveness and robustness of the evaluation are improved.

[0054] The adjustable threshold is set, so that the evaluation result can be flexibly adjusted according to specific application scenarios, and the applicability and practicability of the method are enhanced. BRIEF DESCRIPTION OF DRAWINGS

[0055] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced below. Obviously, the drawings in the following description only constitute some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor.

[0056] Figure 1 A step flow chart of a method for dynamically evaluating endothelial barrier integrity of an organ chip according to the present application;

[0057] Figure 2 A structure block diagram of a system for dynamically evaluating endothelial barrier integrity of an organ chip according to the present application. DETAILED DESCRIPTION

[0058] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined object, the following describes the specific implementation, structure, features and effects of a method for dynamically evaluating endothelial barrier integrity of an organ chip according to the present application in combination with the drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0059] 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 the present application belongs.

[0060] The specific scheme of the method for dynamically evaluating endothelial barrier integrity of an organ chip provided by the present application is specifically described below in combination with the drawings.

[0061] Please refer to Figure 1 which shows a step flow chart of a method for dynamically evaluating endothelial barrier integrity of an organ chip according to the first object of the present application. The method comprises the following steps:

[0062] Step S001: using a confocal microscope to collect high-resolution images of endothelial barrier cells under a fluorescence channel and a transmitted light channel respectively, double-labeled staining of intercellular junction proteins and cell nuclei, and marking cell nucleus region pixel points and intercellular junction protein region pixel points, and obtaining fluorescence images of simultaneously labeled cell nucleus and intercellular junction protein region pixel points.

[0063] In order to realize the evaluation of the endothelial barrier integrity of the organ chip, specific endothelial barrier cell images are first needed as a judgment basis.

[0064] Specifically, high-resolution images of the endothelial barrier cells are collected under a fluorescence channel and a transmission light channel respectively by using a confocal microscope, intercellular junction proteins and cell nuclei are double-labeled and marked cell nucleus region pixel points and intercellular junction protein region pixel points, fluorescence images of the simultaneously labeled cell nucleus and intercellular junction protein region pixel points are obtained, and the specific method is as follows:

[0065] The endothelial barrier cells are double-labeled by using an antibody for specific intercellular junction proteins and DAPI nuclear dye, and then high-resolution images are collected under a fluorescence channel and a transmission light channel respectively by using a confocal microscope, and all shooting parameters are ensured to be consistent. The size of the image is 1080*1080, the region with fluorescence reaction in the gray scale image scanned by 405nm laser is marked as the cell nucleus region, the region with fluorescence reaction in the gray scale image scanned by 488nm laser is marked as the intercellular protein region, and the transmission light image of the endothelial barrier cells and the fluorescence image of the endothelial barrier cells are obtained.

[0066] It should be noted that the image size in this embodiment is 1080*1080, and the image size is not specifically limited by the present application, but the area occupied by the endothelial barrier in the image needs to be at least more than ninety percent of the image area.

[0067] Step S002: The pixel points of the transmission light image of the endothelial barrier cells are clustered into a high gray value class and a low gray value class according to the gray value of the transmission light image of the endothelial barrier cells, and the cell coverage rate W is obtained according to the ratio of the number of pixel points in the low gray value class to the number of all pixel points in the transmission light image of the endothelial barrier cells.

[0068] It should be noted that the integrity of the endothelial barrier is mainly determined by the continuity of the intercellular junction structure and the protein expression intensity, and these microscopic morphological characteristics can be directly presented and quantified in the image by specific fluorescence staining. Therefore, by analyzing the cell morphology, distribution density and fluorescence signal of the intercellular junction proteins, the integrity of the barrier function can be evaluated non-invasively and with high precision, so as to realize the reliable inference of the physiological state.

[0069] ​It is further needed to be explained that only when it is confirmed that the cells have formed a continuous and high-coverage monolayer structure, the barrier failure caused by cell loss can be effectively excluded, thereby avoiding misleading the microscopic analysis of the uniformity of subsequent cell distribution and the strength of connexin function. In the transmission light imaging, the absorption and scattering of light by the cell structure can significantly reduce the transmission light intensity, resulting in a lower gray value of the corresponding imaging area. On the contrary, the blank area without cell coverage has good light transmission, and the transmission light is almost not attenuated, so it shows a high gray value. In view of this gray difference, the image is divided into cell coverage area and non-cell coverage area, and then the coverage degree of the cell layer is accurately evaluated by calculating the area ratio.

[0070] Specifically, the pixel points of the endothelial barrier cell transmission light image are clustered into a high gray value class and a low gray value class according to the gray value of the endothelial barrier cell transmission light image, and the cell coverage rate W is obtained according to the ratio of the number of pixel points in the low gray value class to the number of all pixel points in the endothelial barrier cell transmission light image. The specific method is as follows:

[0071] The gray values of the pixel points in the endothelial barrier cell transmission light image are counted, and the maximum and minimum values among them are taken as the initial cluster centers. Each pixel point is divided into the category corresponding to the cluster center with which the gray value is closer. The gray mean value of all pixel points in each category is recalculated and updated as a new cluster center. The pixel division and cluster center updating process is repeated until the iteration termination condition is met: after a certain iteration, the change amount of the gray values of the two newly calculated cluster centers compared with the previous iteration is less than 1% of the current cluster center gray value. Based on the last pixel division result, all pixel points in the image are clustered into two categories of high gray value and low gray value. The ratio of the number of pixel points in the low gray value category to the number of pixel points in the entire endothelial barrier cell transmission light image is taken as the cell coverage rate W. If the ratio of the number of pixel points in the category with fewer pixel points to the number of pixel points in the entire endothelial barrier cell transmission light image is less than or equal to 5%, the cell coverage rate W = 1.

[0072] It is needed to be explained that this step realizes the automatic identification and quantification of the cell coverage area. The core lies in the inherent difference in light transmission between the cell structure and the background area: the cell area will significantly reduce the transmission light intensity due to the absorption and scattering of light, showing a lower gray value; while the blank area without cells has good light transmission, showing a higher gray value. Based on this physical property, this method uses an unsupervised clustering algorithm to automatically classify the pixel gray value of the transmission light image, and divides the image into cell coverage area and non-cell coverage area.

[0073] Step S003: detecting the position and pixel point number of the nucleus region in the fluorescence image, obtaining the nucleus regions having adjacent relationship with each nucleus region by using Delaunay triangulation, taking the maximum distance of all adjacent nucleus regions as the side length of the window, dividing the image into N windows, and obtaining the nucleus distribution uniformity parameter according to the information entropy of the number of nucleus regions in different windows.

[0074] It should be noted that the integrity of the endothelial barrier fundamentally depends on the order of the cell monolayer in the spatial structure, and therefore the uniformity of the cell distribution is a key basis for evaluating the barrier function. A uniformly spread cell layer ensures continuous physical coverage and avoids innate permeability defects caused by local sparseness; at the same time, such uniformity in structure is also the basis for the formation of stable mechanical coupling between intercellular connection proteins, enabling the barrier to uniformly disperse external stress and maintain overall stability. In theory, all cell nuclei are uniformly distributed in the cell layer, and the uniformity changes due to the influence of external factors. The position uncertainty of the distribution is now calculated as a basis for judging the uniformity of the distribution.

[0075] Specifically, the position and pixel point number of the nucleus region in the fluorescence image are detected, the nucleus regions having adjacent relationship with each nucleus region are obtained by using Delaunay triangulation, the maximum distance of all adjacent nucleus regions is taken as the side length of the window, the image is divided into N windows, and the nucleus distribution uniformity parameter is obtained according to the information entropy of the number of nucleus regions in different windows. The specific method is as follows:

[0076] In the fluorescence image, the geometric center of each pixel point connected domain marked as a nucleus region is calculated as the center point of each nucleus. The Delaunay triangulation algorithm is applied, a triangular grid is constructed based on the coordinates of all nucleus center points, and the adjacent relationship between the nucleus center pairs is determined according to the connection relationship formed by the triangulation to obtain all pixel point connected domains marked as nucleus regions having adjacent relationship. The maximum distance of all nucleus center pairs is taken as the side length of the analysis window, and the entire fluorescence image is divided into windows according to the side length of the window. The number of nucleus region pixel point connected domains in each window is recorded, and the nucleus distribution uniformity parameter is obtained by the normalized entropy of the number distribution of nucleus region pixel point connected domains in different windows. The specific method is as follows:

[0077]

[0078] In the formula, Unif represents the nucleus distribution uniformity parameter, N represents the number of windows, represents the total number of nucleus region pixel point connected domains, and represents the total number of nucleus region pixel point connected domains, and ​The number of pixel points in the nuclear region in the first window is represented as N1. The number of pixel points in the nuclear region in the first window is represented as N1.

[0079] It should be noted that the uniformity of the endothelial cell layer is quantified by analyzing the spatial distribution of the nuclei, which is a key structural indicator for evaluating barrier integrity. This method objectively defines the spatial adjacency relationship between cells using Delaunay triangulation, ensuring the accuracy of adjacency determination and avoiding the subjectivity of manual partitioning. Using the farthest adjacency distance as the window size has the advantage of being adaptive to the current distribution density of the cells, ensuring that each analysis window covers a typical local area, making the analysis results of different samples or different density conditions comparable. Finally, the uniformity is evaluated by calculating the information entropy of the number distribution of cells in each window and normalizing it. The greater the information entropy, the more random and non-uniform the distribution of cells in different windows; after standardization and negative processing, the greater the value of the parameter p1, the more uniform the distribution of cells. This parameter can sensitively capture the distribution heterogeneity caused by cell aggregation or sparseness, and such structural disorder is often an early sign of impaired barrier function.

[0080] Step S004: Select the non-nuclear region between the nearest nuclear region connected domain with adjacent relationship to each nuclear region, and record this region as the connexin template region between the two adjacent nuclei. Take the mean value of the gray values of all pixel points marked as connexin corresponding pixel points in the connexin template region as the intercellular connexin intensity of the two adjacent nuclei. Calculate the Pearson correlation coefficient of the intercellular distance and the intercellular connexin intensity and normalize it to obtain the intercellular connexin intensity parameter.

[0081] It should be noted that the uniformity of the structure only reflects the spatial arrangement of the cells, while the function of the barrier essentially depends on the effective adhesion formed by the intercellular connexin. The increase in nuclear distance means that the cells are actively shrinking or passively pulling apart, which belongs to structural separation; the connexin intensity represents the functional compensation made by the cells to maintain the connection. The negative correlation between the two indicates that the cells are resisting external forces and trying to maintain the continuity of the barrier - this is a healthy stress response, proving the self-stabilizing ability of the barrier.

[0082] Specifically, select the non-nuclear region between the nearest nuclear region connected domain with adjacent relationship to each nuclear region, and record this region as the connexin template region between the two adjacent nuclei. Take the mean value of the gray values of all pixel points marked as connexin corresponding pixel points in the connexin template region as the intercellular connexin intensity of the two adjacent nuclei. Calculate the Pearson correlation coefficient of the intercellular distance and the intercellular connexin intensity and normalize it to obtain the intercellular connexin intensity parameter, including the specific method:

[0083] The geometric center of the nearest nuclear region connected with each nuclear region having an adjacent relationship is connected as a reference line, and a series of parallel lines are generated by pixel translation in the vertical direction to both sides of the reference line until the newly generated parallel line no longer intersects with two nuclear region connected domains, the first intersection point passing through two nuclei and two nuclear region connected domains is recorded, and the pixel point of the line segment between the two points is recorded as the effective connection region on the parallel line. All pixel points on the effective connection region are collected as the connecting protein template region between two adjacent nuclei, and the mean of the gray values of all pixel points marked as connecting protein corresponding pixels in the connecting protein template region is taken as the intercellular connecting protein intensity of two adjacent nuclei. The Pearson correlation coefficient of the intercellular distance and the intercellular connecting protein intensity is calculated and normalized to obtain the intercellular connecting protein intensity parameter. The specific method is as follows:

[0084] ;

[0085]

[0086] In the formula, represents the linear relationship between the nuclear distance and the intercellular connecting protein intensity, represents the total number of nuclear region pixel point connected domains, represents the geometric center distance between the i-th nuclear region pixel point connected domain and the nearest nuclear pixel point connected domain, represents the intercellular protein intensity between the i-th nuclear region pixel point connected domain and the nearest nuclear pixel point connected domain, represents the mean of the geometric center distance between each nuclear region pixel point connected domain and the nearest nuclear region pixel point connected domain, represents the mean of the intercellular protein intensity between each nuclear region pixel point connected domain and the nearest nuclear pixel point connected domain, represents the intercellular connecting protein intensity parameter.

[0087] It should be noted that this step evaluates the degree of functional integrity of the endothelial barrier, not just the static structural existence, by quantifying the statistical relationship between the nuclear distance and the corresponding connexin intensity to reveal the functional compensation ability of cells in maintaining barrier integrity when facing external stress or stimulation. The advantages are: first, by generating the "connexin template area", the effective area to be analyzed between each pair of adjacent cells is accurately defined, avoiding the inclusion of the entire intercellular region or irrelevant background into the calculation, ensuring the accuracy and specificity of the data; second, the Pearson correlation coefficient p is used to objectively measure the linear relationship between structural separation (distance increase) and functional compensation (connexin intensity change). Formula logic: a healthy and elastic barrier will compensate by increasing the expression or aggregation of connexins when the cells are slightly pulled apart, showing a negative correlation between distance and intensity (p is negative). By normalizing the transformation , the correlation coefficient p is mapped to the interval [0, 1], and the ideal negative correlation (p→-1) corresponds to a higher parameter value (→1), so that the barrier function is intuitively represented as good; on the contrary, positive correlation or irrelevant (p≥0) indicates that the barrier function is out of balance or loses compensatory ability, corresponding to a lower p_2 value. This provides a dynamic and sensitive functional strength indicator.

[0088] Step S005: Set up an image grid, filter all grids with labeled intercellular connexin pixel points, and take the pixel point with the lowest gray value labeled as intercellular connexin pixel point in each grid as the seed point. According to the gray value size of the seed point eight-neighborhood pixel points as the growth condition and update the seed point position, find the suspected weak area connected domain in each grid, and according to the ratio of the number of pixel points in the suspected weak area connected domain in each grid to the number of pixel points in the grid, obtain the intercellular connexin integrity parameter.

[0089] It should be noted that the local integrity analysis realizes the spatial resolution of the distribution, continuity and signal intensity of intercellular connexin by fine quantization at the subunit level of each cell connection, so as to effectively identify the microscopic structural weak links that cannot be reflected by statistical average. This microscopic scale can be verified and supplemented by the macroscopic conclusion based on overall correlation, making up for the bias that may be caused by simply relying on statistical trends, and finally providing a more comprehensive, reliable and spatially resolved basis for the integrity of endothelial barrier function.

[0090] ​Specifically, an image grid is set, all grids in which intercellular junction protein pixel points exist are screened, the pixel point with the lowest gray value of the intercellular junction protein pixel points in each grid is taken as a seed point, the gray value of the eight-neighborhood pixel points of the seed point is taken as a growth condition and the position of the seed point is updated, the suspected weak area connected domain in each grid is found, and the intercellular junction protein integrity parameter is obtained according to the ratio of the number of pixel points in the suspected weak area connected domain in each grid to the number of pixel points in the grid, and the specific method comprises:

[0091] An image grid is set in the fluorescence image, the fluorescence image is evenly divided into grids with a size of All grids covering intercellular junction protein pixel points are recorded and are referred to as intercellular protein grids, the mean value of the gray values of all intercellular junction protein pixel points in the yth grid is calculated The intercellular protein pixel point with the lowest gray value in each intercellular protein grid is taken as the first pixel point of the suspected weak area, and if there is a pixel point with a gray value lower than in the eight-neighborhood intercellular protein pixel points of the first pixel point, the pixel point with a gray value lower than is marked as a suspected weak area pixel point in the grid, and it is determined whether there is a pixel point with a gray value lower than in the eight-neighborhood intercellular protein pixel points of the newly marked suspected weak area pixel point, if there is a pixel point with a gray value lower than , the pixel point with a gray value lower than is marked as a suspected weak area pixel point in the grid, and the above eight-neighborhood judgment and suspected weak area pixel point marking process is repeated until all pixel points in the grid are judged, and the specific method for calculating the intercellular junction protein integrity parameter is as follows:

[0092]

[0093] In the formula, represents the intercellular junction protein integrity parameter, represents the total number of grids covering intercellular junction protein pixel points, represents the number of suspected weak area pixel points in the nth grid, represents the number of intercellular junction protein pixel points in the nth grid.

[0094] ​​​It should be noted that the size of the grid is not specifically limited in the present application, and the purpose of selecting the size of the grid is to reduce the influence of noise. In the present embodiment, the size of the grid is 108*108 pixels. The present step aims to actively detect the local structural defects of the connecting protein network in the endothelial barrier from a microscopic scale. These defects may be hidden in the overall average intensity analysis, but they are the potential starting point of barrier dysfunction. The method advantage is that the global image is decomposed into local analysis units through grid processing, ensuring that the entire barrier area is scanned without omission. Then, a region growing-based algorithm is used, taking the average gray value of the connecting protein pixels in each grid as the dynamic threshold, starting from the darkest pixel point, and searching and marking all continuously distributed low-intensity pixel regions (i.e. suspected weak regions). This method can accurately identify small areas of insufficient or broken protein expression, rather than just relying on the overall gray mean value. The formula logic is as follows: first, calculate the proportion of weak area pixels in each grid The higher the ratio, the worse the local integrity of the grid. Then, the average value of the ratio of all grids is obtained to get the overall weakness degree. Finally, subtract the overall weakness degree from 1 to convert the index into an intuitive integrity parameter, so The value of the closer to 1, the better the overall continuity of the connecting protein network, and the fewer the local weak points. On the contrary, the lower the value, the more micro-defects exist in the barrier, and the poorer the integrity.

[0095] Step S006: Obtain the endothelial barrier integrity probability parameter by combining the cell coverage, the cell nucleus distribution uniformity parameter, the intercellular connecting protein intensity parameter, and the intercellular connecting protein integrity parameter.

[0096] It should be noted that the integrity of the endothelial barrier is a multi-dimensional comprehensive attribute, which depends on the formation of continuous coverage by a sufficient number of cells, requires the ordered spatial arrangement of cells, and is more dependent on the functional strength and structural continuity of the intercellular connecting protein. The absence or weakness of any one aspect may lead to impaired barrier function, so the present step combines the evaluation of the real state of the barrier by the above-mentioned step parameters.

[0097] Specifically, the endothelial barrier integrity probability parameter is obtained by combining the cell coverage, the cell nucleus distribution uniformity parameter, the intercellular connecting protein intensity parameter, and the intercellular connecting protein integrity parameter. The specific method includes:

[0098]

[0099] In the formula, P represents the endothelial barrier integrity probability parameter, W represents the cell coverage, represents the cell nucleus distribution uniformity parameter, represents the intercellular connecting protein intensity parameter, represents the intercellular connecting protein integrity parameter.

[0100] It should be noted that the multiplication operation can sensitively amplify the defects of any sub-item, all parameters are normalized to the interval [0, 1], so that the P value itself also becomes a probability index between 0 and 1, and the higher the value represents the more complete the barrier, and the evaluation result is intuitive and reliable. This integrated method greatly improves the robustness and accuracy of the evaluation, avoids the misjudgment that may be caused by relying on a single indicator, and provides a reliable and quantifiable decision basis for subsequent barrier integrity judgment.

[0101] Step S007: Set the endothelial barrier integrity probability parameter threshold, compare the endothelial barrier integrity probability parameter with the endothelial barrier integrity probability parameter threshold, and judge whether the endothelial barrier is complete.

[0102] It should be noted that this step converts the continuous probability parameter P into a clear binary judgment by setting a clear threshold ω. The advantage is its flexibility and interpretability. Users can customize the threshold ω according to the application scenario of the specific organ chip, such as the blood-brain barrier requiring high integrity, while other barriers allowing lower permeability, or the specific requirements of the experiment, so that the evaluation standard is closely linked to the biological function requirements. The threshold is obtained by collecting a large number of sample images with known states based on historical data or control experiments, calculating their P values and drawing a distribution graph, and finding the best classification state P value critical point as the threshold. In this embodiment, the threshold value is 0.16, and in other embodiments, the threshold value is determined according to the specific implementation.

[0103] Specifically, the endothelial barrier integrity probability parameter threshold is set, and the endothelial barrier integrity probability parameter is compared with the endothelial barrier integrity probability parameter threshold to judge whether the endothelial barrier is complete. The specific method is:

[0104] According to the specific organ chip application, the corresponding threshold is set As a specific requirement for evaluating endothelial barrier integrity, if the endothelial barrier integrity probability parameter is greater than the set threshold , it is considered that the endothelial barrier integrity is good, and if the endothelial barrier integrity probability parameter is less than or equal to the set threshold , it is considered that the endothelial barrier integrity is poor.

[0105] Please refer to Figure 2 which shows the structure block diagram of the second object of the present application, a dynamic evaluation system for endothelial barrier integrity of an organ chip. The system includes the following modules:

[0106] The data acquisition module is configured to acquire high-resolution images of endothelial barrier cells under a fluorescence channel and a transmission light channel respectively by using a confocal microscope, to perform double-labeling staining on an intercellular junction protein and a cell nucleus, to mark pixel points in a cell nucleus region and pixel points in an intercellular junction protein region, and to acquire a fluorescence image in which the cell nucleus and the intercellular junction protein are simultaneously labeled.

[0107] The cell coverage calculation module is configured to cluster pixel points in the transmission light image of the endothelial barrier cells into a high gray value class and a low gray value class according to gray values of the transmission light image of the endothelial barrier cells, and to acquire the cell coverage W according to a ratio of a number of pixel points in the low gray value class to a total number of pixel points in the transmission light image of the endothelial barrier cells.

[0108] The cell nucleus distribution uniformity parameter calculation module is configured to detect positions and numbers of pixel points of cell nucleus regions in the fluorescence image, to acquire cell nucleus regions having an adjacent relationship with each cell nucleus region by using a Delaunay triangular net, to divide the image into N windows by taking a farthest distance of all adjacent cell nucleus regions as a window side length, and to acquire a cell nucleus distribution uniformity parameter according to an information entropy of numbers of cell nucleus regions in different windows.

[0109] The intercellular junction protein intensity parameter calculation module is configured to select a non-cell nucleus region between a nearest cell nucleus region connected domain having an adjacent relationship with each cell nucleus region, to record the region as a junction protein template region between two adjacent cell nuclei, to take a mean value of gray values of all pixel points marked as junction protein corresponding pixel points in the junction protein template region as an intercellular junction protein intensity of the two adjacent cell nuclei, to acquire a Pearson correlation coefficient of the intercellular distance and the intercellular junction protein intensity and normalize the Pearson correlation coefficient, and to acquire an intercellular junction protein intensity parameter.

[0110] The intercellular junction protein integrity parameter calculation module is configured to screen all grids in which there are pixel points marked as intercellular junction protein pixel points, to take a pixel point with a lowest gray value of the intercellular junction protein pixel points in each grid as a seed point, to find a suspected weak region connected domain in each grid according to gray values of eight-neighborhood pixel points of the seed point as a growth condition and update a position of the seed point, and to acquire an intercellular junction protein integrity parameter according to a ratio of a number of pixel points in the suspected weak region connected domain in each grid to a number of pixel points in the grid.

[0111] The endothelial barrier integrity probability parameter calculation module is configured to acquire an endothelial barrier integrity probability parameter by combining the cell coverage, the cell nucleus distribution uniformity parameter, the intercellular junction protein intensity parameter, and the intercellular junction protein integrity parameter.

[0112] The endothelial barrier integrity judgment module is configured to set an endothelial barrier integrity probability parameter threshold value, and compare the endothelial barrier integrity probability parameter with the endothelial barrier integrity probability parameter threshold value to determine whether the endothelial barrier is complete.

[0113] The present application has the following advantages:

[0114] Through double-channel fluorescence imaging and specific wavelength labeling, the accurate distinction between the nucleus and the connexin is ensured, and reliable data basis is provided for subsequent analysis.

[0115] The adaptive clustering method is used to distinguish cells and background, effectively excluding noise interference, and improving the accuracy of cell coverage calculation.

[0116] The Delaunay triangulation and information entropy are combined to objectively quantify the uniformity of the nucleus distribution, avoid the subjectivity of manual division, and enhance the reliability of the analysis.

[0117] By constructing the connexin template region and calculating the distance-intensity correlation, the functional compensation ability of the intercellular connection is revealed, and a dynamic functional evaluation index is provided.

[0118] The gridding and region growing algorithm are used to actively identify local weak areas, make up for the shortcomings of overall average analysis, and improve the sensitivity of defect detection.

[0119] The barrier state is comprehensively evaluated through multi-parameter fusion, avoiding the limitation of a single index, and improving the comprehensiveness and robustness of the evaluation.

[0120] The adjustable threshold value is set, so that the evaluation result can be flexibly adjusted according to the specific application scene, and the applicability and practicability of the method are enhanced.

[0121] Those skilled in the art will appreciate that embodiments of the present application can be provided as methods, systems, or computer program products. Accordingly, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, etc.) having computer-usable program code embodied therein.

[0122] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flow or blocks Figure 1 one or more flow or blocks

[0123] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flow or blocks Figure 1 one or more flow or blocks

[0124] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flow or blocks Figure 1 one or more flow or blocks

[0125] Finally, it should be noted that the above-mentioned embodiments are merely used to illustrate the technical solutions of the present application, rather than limit the technical solutions of the present application. Although the present application is described in detail with reference to the above embodiments, those skilled in the art should understand that the specific embodiments of the present application can be modified or replaced, and any modification or replacement without departing from the spirit and scope of the present application should be covered in the protection scope of the claims of the present application.

Claims

1. A method for dynamic assessment of endothelial barrier integrity on an organ-on-a-chip, characterized in that, The method includes the following steps: High-resolution images of endothelial barrier cells were acquired using a confocal microscope under both the fluorescence channel and the transmission light channel. Dual labeling of intercellular junction proteins and cell nuclei was performed, and pixels in the cell nucleus region and the intercellular junction protein region were labeled to obtain fluorescence images that simultaneously labeled pixels in the cell nucleus and intercellular junction protein regions. Based on the gray values ​​of the translucent images of endothelial barrier cells, the pixels in the translucent images of endothelial barrier cells are clustered into high gray value classes and low gray value classes. The cell coverage W is obtained by the ratio of the number of pixels in the low gray value class to the total number of pixels in the translucent images of endothelial barrier cells. The location and number of pixels of the cell nucleus region in the fluorescence image are detected. The Delaunay triangulation is used to obtain the cell nucleus regions that are adjacent to each cell nucleus region. The image is divided into N windows with the farthest distance of all adjacent cell nucleus regions as the window side length. The cell nucleus distribution uniformity parameter is obtained based on the information entropy of the number of cell nucleus regions in different windows. The non-nuclear region between the nearest connected domain of each nucleus region is selected and recorded as the connective protein template region between two adjacent nuclei. The mean gray value of all pixels marked as connective proteins in the connective protein template region is used as the intercellular connective protein strength between two adjacent nuclei. The Pearson correlation coefficient between intercellular distance and intercellular connective protein strength is calculated and normalized to obtain the intercellular connective protein strength parameter. Set up an image grid, filter all grids containing pixels marked as intercellular connective proteins, use the pixel with the lowest gray value among the pixels marked as intercellular connective proteins in each grid as the seed point, use the gray values ​​of the eight neighboring pixels of the seed point as the growth condition and update the position of the seed point, find the suspected weak region connected domain in each grid, and obtain the intercellular connective protein integrity parameter based on the ratio of the number of pixels in the suspected weak region connected domain to the number of pixels in the grid. The probability parameters of endothelial barrier integrity are obtained by combining parameters such as cell coverage, uniformity of cell nuclear distribution, strength of intercellular junction proteins, and integrity of intercellular junction proteins. Set a threshold for the probability parameter of endothelial barrier integrity, and compare the probability parameter of endothelial barrier integrity with the threshold to determine whether the endothelial barrier is intact.

2. The method for dynamic assessment of endothelial barrier integrity on an organ-on-a-chip according to claim 1, characterized in that, The method for acquiring high-resolution images of endothelial barrier cells using a confocal microscope under both fluorescence and transmission light channels, performing dual-labeling staining on intercellular junction proteins and the cell nucleus, and labeling pixels in both the nuclear and intercellular junction protein regions to obtain fluorescence images simultaneously labeled with pixels in both regions, includes the following specific steps: Endothelial barrier cells were double-labeled using antibodies targeting specific intercellular junction proteins and DAPI nuclear dye. High-resolution images were then acquired using confocal microscopy under both fluorescence and transmission light channels, ensuring all imaging parameters remained consistent. The image size was [missing information]. The images were analyzed, and the areas showing fluorescence in the grayscale image scanned by the 405nm laser were marked as the cell nucleus region, and the areas showing fluorescence in the grayscale image scanned by the 488nm laser were marked as the intercellular protein region, thus obtaining the transluminal image and fluorescence image of the endothelial barrier cells.

3. The method for dynamic assessment of endothelial barrier integrity on an organ-on-a-chip according to claim 1, characterized in that, The specific method for clustering pixels in the translucent image of endothelial barrier cells into high gray value classes and low gray value classes based on gray values, and obtaining the cell coverage W based on the ratio of the number of pixels in the low gray value class to the total number of pixels in the translucent image of endothelial barrier cells, includes the following: The grayscale values ​​of each pixel in the translucent image of endothelial barrier cells are statistically analyzed, and the maximum and minimum values ​​are used as the initial cluster centers. Each pixel is assigned to the category corresponding to the cluster center whose grayscale value is closer to its own. The mean grayscale value of all pixels in each category is recalculated and updated as the new cluster center. The above pixel division and cluster center update process is repeated until the iteration termination condition is met: after a certain iteration, the change in the grayscale values ​​of the two newly calculated cluster centers compared with the previous one is less than 1% of the current cluster center's grayscale value. Based on the last pixel division result, all pixels in the image are clustered into two categories: high grayscale value and low grayscale value. The ratio of the number of pixels in the low grayscale value category to the total number of pixels in the translucent image of endothelial barrier cells is used as the cell coverage rate W. If the ratio of the number of pixels in the category with fewer pixels to the total number of pixels in the translucent image of endothelial barrier cells is less than or equal to 5%, the cell coverage rate W = 1.

4. The method for dynamic assessment of endothelial barrier integrity on an organ-on-a-chip according to claim 1, characterized in that, The method involves detecting the location and number of pixels in the cell nucleus region of the fluorescence image, using a Delaunay triangulation to obtain the cell nucleus regions adjacent to each cell nucleus region, dividing the image into N windows with the farthest distance between all adjacent cell nucleus regions as the window side length, and obtaining the cell nucleus distribution uniformity parameter based on the information entropy of the number of cell nucleus regions in different windows. The specific methods include: In the fluorescence image, the geometric center of the connected region of each pixel labeled as a cell nucleus is calculated and used as the center point of each cell nucleus. The Delaunay triangulation algorithm is applied to construct a triangular mesh based on the coordinates of all cell nucleus center points. The adjacency relationship between pairs of cell nucleus centers is determined based on the connectivity relationships formed by this triangulation. All connected regions of pixels labeled as cell nucleus regions with adjacency relationships are obtained, and the maximum distance among all pairs of cell nucleus center points is used as the criterion. The entire fluorescence image is divided according to the side length of the analysis window. The window is segmented, and the number of connected components in the cell nucleus region of each window is recorded. The uniformity parameter of cell nucleus distribution is obtained by the standardized entropy of the distribution of the number of connected components in the cell nucleus region of different windows. The specific method is as follows: In the formula, A parameter representing the uniformity of cell nucleus distribution, where N represents the number of windows. This represents the total number of connected components in the pixel region of the cell nucleus. Indicates the first The number of connected components in the cell nucleus region of each window.

5. The method for dynamic assessment of endothelial barrier integrity on an organ-on-a-chip according to claim 1, characterized in that, The process involves selecting a non-nuclear region adjacent to the nearest connected domain of each nucleus region and designating this region as the connective protein template region between two adjacent nuclei. The average grayscale value of all pixels marked as connective proteins within this template region is used as the intercellular connective protein strength between the two adjacent nuclei. The Pearson correlation coefficient between intercellular distance and intercellular connective protein strength is calculated and normalized to obtain the intercellular connective protein strength parameter. The specific methods include: Using the geometric center of the nearest connected domain of each adjacent nucleus region as a baseline, a series of parallel lines are generated by translating pixel by pixel in the vertical direction on both sides of this baseline, until the newly generated parallel lines no longer intersect with the connected domains of two nucleus regions simultaneously. The first intersection point that simultaneously passes through two nuclei and intersects with the connected domains of two nucleus regions is recorded. The pixel containing the line segment between the two points is recorded as the effective connection region on the parallel line. The set of all pixels on all effective connection regions is recorded as the connector protein template region between two adjacent nuclei. The average gray value of all pixels marked as connector proteins in the connector protein template region is used as the intercellular connector protein strength between two adjacent nuclei. The Pearson correlation coefficient between intercellular distance and intercellular connector protein strength is calculated and normalized. The specific method for obtaining the intercellular connector protein strength parameter is as follows: ； In the formula, This represents the linear relationship between internuclear distance and the strength of intercellular junction proteins. This represents the total number of connected components in the pixel region of the cell nucleus. This represents the distance between the connected regions of the i-th cell nucleus pixel and the geometric center of the connected region of its nearest neighbor cell nucleus pixel. This represents the intercellular protein strength between the connected domain of the i-th cell nucleus pixel and the connected domain of its nearest neighboring cell nucleus pixel. This represents the mean distance between the geometric centers of the connected regions of each pixel in the cell nucleus region and the connected regions of its nearest neighboring pixel in the cell nucleus region. This represents the average intercellular protein intensity between the connected regions of each pixel in the cell nucleus region and the connected regions of its nearest neighboring cell nucleus pixel. This represents the strength parameter of intercellular junction proteins.

6. The method for dynamic assessment of endothelial barrier integrity on an organ-on-a-chip according to claim 1, characterized in that, The process of setting up an image grid involves filtering all grids that contain pixels marked as intercellular connective proteins. The pixel with the lowest grayscale value among these labeled pixels in each grid is designated as the seed point. The seed point's position is updated based on the grayscale values ​​of its eight neighboring pixels. This process identifies potential weak regions within each grid and obtains the intercellular connective protein integrity parameter based on the ratio of the number of pixels in these potential weak regions to the total number of pixels in the grid. The specific methods include: Set an image grid in the fluorescence image to divide the fluorescence image into equal parts of size. A grid is created that records all grid points covering intercellular connection protein pixels, denoted as the intercellular protein grid. The total number of pixels within the y-th grid is calculated. Average gray value of each pixel of intercellular connection protein The intercellular protein pixel with the lowest gray value in each intercellular protein grid is recorded as the first pixel of the suspected weak region. Within the eight-neighborhood of the first pixel, if any intercellular protein pixel has a gray value lower than... For pixels with a grayscale value lower than 0, the corresponding pixel will be classified as a grayscale value lower than 0. The pixels marked as suspected weak areas in the grid are then analyzed to determine whether there are any intercellular protein pixels with a gray value lower than the eight neighboring pixels of the newly marked suspected weak areas. If there are pixels with gray values ​​lower than 1000, For pixels with a grayscale value lower than 0, the corresponding pixel will be classified as a grayscale value lower than 0. The pixels marked as suspected weak areas in the grid are repeated. The process of eight-neighbor judgment and marking suspected weak areas is repeated until all pixels in the grid have been judged. The specific method for calculating the integrity parameters of intercellular connection proteins is as follows: In the formula, Parameters representing the integrity of intercellular junction proteins. Indicates a total of A grid covering pixels of intercellular connection proteins. Indicates the first The number of pixels in suspected weak areas within each grid. Indicates the first Within each grid Individual intercellular connection protein pixels.

7. The method for dynamic assessment of endothelial barrier integrity on an organ-on-a-chip according to claim 1, characterized in that, The specific method for obtaining the endothelial barrier integrity probability parameter by combining parameters such as cell coverage, uniformity of cell nuclear distribution, strength of intercellular junction proteins, and integrity of intercellular junction proteins includes: In the formula, P represents the probability parameter of endothelial barrier integrity, and W represents the cell coverage. A parameter representing the uniformity of cell nucleus distribution. This represents a parameter indicating the strength of intercellular junction proteins. This represents a parameter indicating the integrity of intercellular junction proteins.

8. The method for dynamic assessment of endothelial barrier integrity on an organ-on-a-chip according to claim 1, characterized in that, The specific method for determining whether the endothelial barrier is intact by setting a threshold for the endothelial barrier integrity probability parameter and comparing the endothelial barrier integrity probability parameter with the threshold is as follows: Set appropriate thresholds based on the specific uses of organ-on-a-chip. As a specific requirement for assessing endothelial barrier integrity, if the endothelial barrier integrity probability parameter Greater than the set threshold If the endothelial barrier integrity is considered good, then the probability parameter of endothelial barrier integrity is considered to be... Less than or equal to the set threshold This indicates poor endothelial barrier integrity.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method for dynamic assessment of organ-on-a-chip endothelial barrier integrity as described in any one of claims 1 to 8.

10. An organ-on-a-chip dynamic assessment system for endothelial barrier integrity, characterized in that, The system includes the following modules: The data acquisition module is used to acquire high-resolution images of endothelial barrier cells under both the fluorescence channel and the transmission light channel using a confocal microscope. It performs dual labeling staining on intercellular junction proteins and cell nuclei, and labels the pixels in the cell nucleus region and the pixels in the intercellular junction protein region, and obtains fluorescence images that simultaneously label the pixels in the cell nucleus and the intercellular junction protein region. The cell coverage calculation module is used to cluster the pixels of the endothelial barrier cell transparent image into high gray value class and low gray value class according to the gray value of the endothelial barrier cell transparent image, and obtain the cell coverage W according to the ratio of the number of pixels in the low gray value class to the total number of pixels in the endothelial barrier cell transparent image. The cell nucleus distribution uniformity parameter calculation module is used to detect the position and number of pixels of the cell nucleus region in the fluorescence image. It uses Delaunay triangulation to obtain the cell nucleus regions that are adjacent to each cell nucleus region. The image is divided into N windows with the farthest distance of all adjacent cell nucleus regions as the window side length. The cell nucleus distribution uniformity parameter is obtained based on the information entropy of the number of cell nucleus regions in different windows. The intercellular connective protein strength parameter calculation module is used to select the non-nuclear region between the nearest connected domain of the nuclear region that is adjacent to each nuclear region, and record this region as the connective protein template region between two adjacent nuclei. The average gray value of all pixels marked as connective proteins in the connective protein template region is used as the intercellular connective protein strength between two adjacent nuclei. The Pearson correlation coefficient between intercellular distance and intercellular connective protein strength is calculated and normalized to obtain the intercellular connective protein strength parameter. The intercellular connective protein integrity parameter calculation module is used to filter all grids containing pixels marked as intercellular connective proteins. The pixel with the lowest gray value among the pixels marked as intercellular connective proteins in each grid is used as the seed point. The gray values ​​of the eight neighboring pixels of the seed point are used as the growth condition and the position of the seed point is updated. The module finds the suspected weak region connected domains in each grid and obtains the intercellular connective protein integrity parameter based on the ratio of the number of pixels in the suspected weak region connected domains to the number of pixels in the grid. The endothelial barrier integrity probability parameter calculation module is used to combine cell coverage, cell nucleus distribution uniformity parameters, intercellular junction protein strength parameters, and intercellular junction protein integrity parameters to obtain endothelial barrier integrity probability parameters. The endothelial barrier integrity assessment module is used to set a threshold for the endothelial barrier integrity probability parameter and compare the endothelial barrier integrity probability parameter with the threshold to determine whether the endothelial barrier is intact.

Images