TWI813145B - Multiple fluorescent tissue image analysis method and system thereof - Google Patents

Abstract

The present invention provides a multiple fluorescent tissue image analysis method and system thereof. The method comprises obtaining an analysis data of a tissue section image; filtering the analysis data with a first parameter; using a second parameter and a third parameter in the filtered analysis data as an x axis and an y axis of a graph respectively; plotting the corresponding values of the second parameter and the third parameter in the graph to present a plurality of mark points in the graph; and defining an x-axis threshold and a y-axis threshold for the graph to distribute the plurality of the mark points into a plurality of quadrants. The present invention can process the data generated from the fluorescent imaging analysis tools of the prior art and can redefine multiple cell markers for a single cell to classify in detail and to output analysis result intuitively with image cytometry.

Description

Multiple fluorescent tissue image analysis method and system

The present invention relates to an image analysis method and its system, in particular to a multiple fluorescent tissue image analysis method and its system that is easy to screen cells to be analyzed and can define multiple cell markers for a single cell.

Fluorescent immunohistochemical staining provides some visual evidence in immunological research. For example, if you want to look for CD8 ⁺ T cells and want to know whether CD8 ⁺ T cells have cytotoxic function, you can use CD8 as the target marker and stain with red fluorescence. In addition, using the green fluorescence of Granzyme B for staining, the location of activated CD8 ⁺ T cells can be clearly seen on the image at the same time.

In order to more clearly judge the functionality of each cell on the tissue section, existing fluorescence image analysis tools (such as the Phenoptics ^TM Tissue Analysis Software system, developed by Akoya Biosciences and represented by Bock Technology Co., Ltd.) provide the ability to present and distinguish more detailed information at once. The multi-cell marker method, for example, provides seven different fluorescent dyes - CD4, CD8, CD56, Granzyme B, Foxp3, PanCK, and Nucleus to provide seven cell markers (maker), and can use artificial intelligence to slice a large number of tissues Images are analyzed in a short time.

However, when defining cells, existing fluorescence image analysis tools (such as inForm® Image Analysis Software in the Phenoptics ^TM Tissue Analysis Software system) can only select main cells as defined targets. When the number of target cells is too small, existing The fluorescence image analysis tool needs to first find a single cell containing multiple fluorescent signals, and the number of cells must be sufficient (for example, greater than or equal to 5). Therefore, it is impossible to accurately define multiple fluorescent cells for all staining results, and only A fluorescent signal can be defined for a single cell. In some special circumstances, such as when fluorescent signals are highly overlapping, there may also be problems such as difficulty in user judgment and software calculation errors. Therefore, existing fluorescence image analysis tools still have room for improvement in defining a variety of cells.

In order to solve the above-mentioned problems of the prior art, one purpose of the present invention is to provide a multiple fluorescent tissue image analysis method, which includes: obtaining analysis data of a tissue slice image with a plurality of cells and plural analysis images corresponding to the analysis data; from the Analyze the data to screen out those cells that have at least one first parameter, and use the second parameter and the third parameter in the screened cells that have the first parameter as the X-axis and Y-axis of a graph respectively; Select one of the plurality of analysis images corresponding to the second parameter as the first image to be analyzed to obtain the threshold value of the X-axis from the first image to be analyzed, and obtain the threshold value of the Select one of the plurality of analyzed images as the second image to be analyzed to obtain the threshold value of the Y-axis from the second image to be analyzed; and draw the values corresponding to the second parameter and the third parameter on the graph , a plurality of mark points corresponding to the first parameter are presented in the graph, and the plurality of mark points are distributed in a plurality of quadrants according to the threshold of the X-axis and the threshold of the Y-axis.

As in the aforementioned multiple fluorescent tissue image analysis method, the first parameter is cell phenotype, and the cell phenotype is CD8 ⁺ T cell or CD4 ⁺ T cell.

As in the aforementioned multiple fluorescent tissue image analysis method, it further includes using the second parameter and the third parameter as the X-axis and Y-axis of the graph respectively, and filtering the plurality of cells having the first parameter. A fourth parameter is used to differentiate, wherein the fourth parameter is the tissue type, and the tissue type is PanCK ⁺ or PanCK ^- .

As in the aforementioned multiple fluorescent tissue image analysis method, the second parameter or the third parameter is a plurality of fluorescent dye types or autofluorescence of the cell nucleus, cytoplasm, cell membrane, entire cells, and the second parameter or the The value corresponding to the third parameter is the signal intensity of multiple fluorescent dyes or the signal intensity of autofluorescence in the nucleus, cytoplasm, cell membrane, and the entire cell.

As in the aforementioned multiple fluorescence tissue image analysis method, the step of obtaining the threshold value of the X-axis from the first image to be analyzed further includes: obtaining a plurality of first click coordinates from the first image to be analyzed, so as to combine the plural The weakest signal strength among the cells corresponding to the first clicked coordinates is used as the threshold of the X-axis.

As in the aforementioned multiple fluorescent tissue image analysis method, the step of obtaining the threshold value of the Y-axis from the second image to be analyzed further includes: obtaining a plurality of second click coordinates from the second image to be analyzed, so as to convert the plurality of second click coordinates into The second point selects the weakest signal strength among the cells corresponding to the coordinates as the threshold of the Y-axis.

As in the aforementioned multiple fluorescent tissue image analysis method, it further includes: using the second parameter and the third parameter in the plurality of cells having the first parameter after screening as the X-axis and Y-axis of the graph respectively. At the same time, the values corresponding to the second parameter and the third parameter are standardized.

For example, the aforementioned multiple fluorescent tissue image analysis method further includes: defining different colors for the marker points in different quadrants, and based on the XY coordinates of the cells corresponding to the plurality of marker points in the tissue slice image, dividing the different colors into The marked points are output into a cell distribution map corresponding to the tissue section image.

For example, the aforementioned multiple fluorescence tissue image analysis method further includes: appending the plurality of marker points and their corresponding colors to the analysis data to output updated analysis data.

Another object of the present invention is to provide a multiple fluorescent tissue image processing system, including: an input module for obtaining analysis data of a tissue slice image with a plurality of cells and plural analysis images corresponding to the analysis data; screening and coordinates Define a module to filter out those cells with at least one first parameter from the analysis data, and use the second parameter and the third parameter in the filtered cells with the first parameter as respectively The X-axis and Y-axis of a graph; the threshold definition module is used to select one of the plurality of analysis images corresponding to the second parameter as the first image to be analyzed, to obtain from the first image to be analyzed The threshold value of the A group is used to draw the values corresponding to the second parameter and the third parameter in the graph, so as to present a plurality of mark points corresponding to the first parameter in the graph, and according to the threshold of the X-axis and the Y The axis threshold distributes the complex marker points into complex quadrants.

As in the aforementioned multiple fluorescent tissue image processing system, the first parameter is cell phenotype, and the cell phenotype is CD8 ⁺ T cell or CD4 ⁺ T cell.

As in the aforementioned multiple fluorescent tissue image processing system, after using the second parameter and the third parameter as the X-axis and Y-axis of the graph respectively, the filtering and coordinate definition module will filter to have the first The plurality of cells of the parameters are further distinguished by a fourth parameter, wherein the fourth parameter is a tissue type, and the tissue type is PanCK ⁺ or PanCK ⁻ .

As in the aforementioned multiple fluorescent tissue image processing system, the second parameter or the third parameter is a plurality of fluorescent dye types or autofluorescence of the cell nucleus, cytoplasm, cell membrane, the entire cell, and the second parameter or the The value corresponding to the third parameter is the signal intensity of multiple fluorescent dyes or the signal intensity of autofluorescence in the nucleus, cytoplasm, cell membrane, and the entire cell.

As in the aforementioned multiple fluorescent tissue image processing system, the threshold definition module first obtains a plurality of first click coordinates from the first image to be analyzed, so as to convert the signals in cells corresponding to the plurality of first click coordinates into The one with the weakest intensity is used as the threshold of the X-axis.

As in the aforementioned multiple fluorescent tissue image processing system, the threshold definition module first obtains a plurality of second click coordinates from the second image to be analyzed, so as to convert the signals in cells corresponding to the plurality of second click coordinates into The one with the weakest intensity is used as the threshold of the Y-axis.

As in the aforementioned multiple fluorescent tissue image processing system, the screening and coordinate definition module uses the second parameter and the third parameter in the plurality of cells that have the first parameter after screening as the X of the graph respectively. axis and the Y-axis, the values corresponding to the second parameter and the third parameter are normalized.

For example, the aforementioned multiple fluorescent tissue image processing system also includes an output module for defining different colors for the marker points in different quadrants, and based on the XY coordinates of the cells corresponding to the plurality of marker points in the tissue slice image. , output the marker points of different colors into a cell distribution map corresponding to the tissue section image, or append the plurality of marker points and their corresponding colors to the analysis data to output an updated analysis data.

Through the multiple fluorescent tissue image processing method and system of the present invention, the cells to be analyzed can be easily screened out, and more intuitive analysis results can be output, such as the proportion of sub-cell groups, using the image cytometry defined in the present invention. The proportion of main cells and the distribution of sub-cell groups in the ROI image can facilitate users to compare the pathological diagrams provided by existing fluorescence image analysis tools to determine whether the definition is correct. In addition, the present invention can also establish a threshold for each fluorescent signal according to the signal intensity, and define cells above and below the threshold as positive and negative respectively, thereby analyzing the multiple fluorescent positive signals of each cell in the graph to achieve accurate analysis. The purpose of defining multiple cell markers for a single cell is to solve the problem in the previous technology that only one fluorescent signal can be defined for a single cell, so as to provide users with a better understanding of each cell. phenotypic status, and can facilitate the analysis of daughter cell types of a certain cell group.

1:Multiple fluorescent tissue image processing system

11:Input module

12: Filtering and coordinate definition module

13: Threshold definition module

14: Graphic drawing module

15:Output module

21:X-axis threshold

22: Y axis threshold

31: The first quadrant

32:Second Quadrant

33: The third quadrant

34:The fourth quadrant

41: The first desire is to analyze images

411: First click to select coordinates

42: The second desire to analyze images

421: Second click coordinates

S1-S5: Steps

Figure 1 is a flow chart of the steps of the multiple fluorescent tissue image analysis method of the present invention.

Figure 2 is a system architecture diagram of the multiple fluorescent tissue image analysis system of the present invention.

3 and 4 are schematic diagrams of the first/second image to be analyzed and the plurality of first/second click coordinates of the present invention.

FIG. 5 is a schematic diagram of a graph with multiple marker points, X-axis thresholds and Y-axis thresholds according to the present invention.

Figure 6 is a schematic diagram of the cell distribution map corresponding to the tissue section image output by the present invention.

The following describes the implementation of the present invention through specific embodiments, and those familiar with the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification, and can also use other different specific embodiments. carry out or apply.

Please refer to Figure 1. In the step S1 of the multiple fluorescent tissue image analysis method of the present invention, analysis data of a tissue slice image with a plurality of cells and plural analysis images corresponding to the analysis data are first obtained. The following will first take software such as Phenoptics ^TM Tissue Analysis Software system as an example to illustrate the generation method, source, format and related experimental procedures of the analysis data and complex analysis images. However, the present invention does not assume that only this software can generate the analysis data and complex analysis images. Limited to analysis data and complex analysis images (for example, you can also use HALO® image analysis platform, Indica Labs software).

First, a tissue wax block made from the patient's specimen is obtained and sliced into sections so that the thickness can be penetrated by light (for example, 5 μm). Then, the tissue sections are stained as required, for example, seven different fluorescent dyes, including CD4, CD8, CD56, Granzyme B, FoxP3, PanCK, and Nucleus, are used to stain different cell markers. Afterwards, the stained tissue sections can be analyzed using the Phenoptics ^TM Tissue Analysis Software system, such as scanning the entire section, selecting regions of interest (ROI), detailed scanning of each ROI, AI training on part of the ROI, and AI training. The parameters after training are applied to all ROIs to perform analysis, generate analysis results and other analysis steps.

Specifically, after scanning the whole film, a tissue section image with multiple cells can be generated. Multiple ROIs can be circled on each tissue section image, and a total of 75% of the tumor area and 25% of the tumor area will be circled on each tissue section image. benign area. Perform detailed scanning and fluorescence imaging of each ROI. Then, 5 to 10 ROIs can be selected for AI training to subdivide the ROI in the tumor area into tumor cells (Tumor), stromal cells (Stroma), and benign areas (Benign). In one embodiment, the presence or absence of PanCK signal can be used to distinguish tumor cells and stromal cells. For example, cells with PanCK signal (PanCK ⁺ ) are tumor cells, and cells without PanCK signal (PanCK ^- ) are stromal cells.

Next, the user defines the cells on the tissue slice image, such as defining what a cell is and the range of the cell nucleus, cytoplasm, and cell membrane. Cells can be defined based on the presence or absence of the DAPI signal generated by the fluorescent dye Nucleus. That is, the presence of DAPI signal means that it is a cell. Based on the DAPI signal, a size range is defined for the cell nucleus. Readings within this range The fluorescent signals detected are defined as staining signals on the cell nucleus. Similarly, the signal generated by the fluorescent dye Granzyme B determines the range of the cytoplasm receiving the signal, and the signals generated by the fluorescent dyes CD4, CD8, and CD56 determine the range of the cell membrane receiving the signal. After determining the nucleus, cytoplasm, and After reading the range of the cell membrane signal, the user can define the cells on the tissue section image. For example, after switching to the CD8 signal, a specific color area on the image will be defined as Cytotoxic T cell. After repeating this step several times, The AI can be trained (e.g., supervised learning) to train a model that can infer the location of other Cytotoxic T cells in other ROIs. In the same way, CD4 Helper T cells and CD 56 NK cells can also use this definition and training method.

After the analysis is completed, the analysis results can be generated, for example, output into a txt file or excel file as the analysis data of the present invention (for example, output by inForm® Image Analysis Software in the Phenoptics ^TM Tissue Analysis Software system), and each column in the data is analyzed All represent the data of one cell, and each column can correspond to multiple fields. The fields can include: the ROI number where each cell is located, the tissue type (such as PanCK ⁺ or PanCK ^- ), and the cell phenotype (such as CD8 ⁺ T cell or CD4 ⁺ T cell, or Cytotoxic T cell defined by CD8, Helper T cell defined by CD4, NK cell defined by CD56, etc.), cell ID, XY coordinates of the cell in the ROI. In addition to the above fields, there are also the following fields: the nucleus, cytoplasm, cell membrane of each cell, multiple fluorescent dye types and autofluorescence of the entire cell. The values in the fields are the signal intensity. In other words, if seven fluorescent dyes are used, each cell will obtain 32 fluorescent data ((n+1)*4 fluorescent data, n is the number of fluorescent dyes). In addition, the analysis results can also output complex analysis images (pathology diagrams). The complex analysis images are classified by different fluorescent dye types and their ROIs. In other words, each analysis image contains the same fluorescent dye type and ROI. All cells in the same ROI (for example, corresponding to the ROI number in the analysis data, the XY coordinates of the cells in the ROI, and the type of fluorescent dye).

After obtaining the analysis data, step S2 can be performed to screen out the cells with the first parameter from the analysis data, and use the second parameter and the third parameter in the cells with the first parameter after screening to As the X-axis and Y-axis of a graph, the first parameter is the cell phenotype, the cell phenotype is CD8 ⁺ T cell or CD4 ⁺ T cell, the second parameter or the third parameter is the cell nucleus, cytoplasm, cell membrane, Multiple fluorescent dye types or autofluorescence of the entire cell, and the value corresponding to the second parameter or the third parameter is the signal intensity of the nucleus, cytoplasm, cell membrane, multiple fluorescent dyes of the entire cell, or autofluorescence. Fluorescent signal strength. For example, if you want to analyze CD8 ⁺ T cells and want to know whether CD8 ⁺ T cells have a cytotoxic function, the first parameter will select the cell phenotype as CD8 ⁺ T cell, and the second or third parameter will select cytoplasmic Granzyme. B, because Granzyme B is a cell marker expressed in the cytoplasm. In another embodiment, FoxP3 in the nucleus can be selected as the X-axis, and Granzyme B in the cytoplasm can be selected as the Y-axis, to further obtain the difference between FoxP3 in the nucleus and Granzyme B in the cytoplasm of each cell obtained after screening. The corresponding value in the field. In addition, after using the second parameter and the third parameter as the X-axis and Y-axis of the graph respectively, the plurality of cells having the first parameter after screening are further distinguished by a fourth parameter, where the fourth parameter is the tissue type. , the tissue type is PanCK ⁺ or ^PanCK- . Specifically, the analysis data can be filtered by selecting only tissue types, only cell phenotypes, or both tissue types and cell phenotypes. In one embodiment, the cell phenotype can be screened first, and then the tissue type can be screened. For example, cell data whose cell phenotype is CD8 ⁺ T cell can be screened first, and then cell data whose tissue type is PanCK ⁺ can be screened out, etc. The present invention is not limited to this, and vice versa.

In one embodiment, while using the second parameter and the third parameter in the plurality of cells with the first parameter after screening as the X-axis and Y-axis of the graph respectively, the corresponding values of the second parameter and the third parameter can be Normalize the values and compress all values to between 0 and 1 so that the fluorescence intensity of each cell marker has a consistent basis. Then proceed to step S3.

In step S3, one of the plurality of analyzed images corresponding to the second parameter is selected as the first image to be analyzed, so as to obtain the threshold value of the X-axis from the first image to be analyzed, and Select one of the plurality of analyzed images corresponding to the third parameter as the second image to be analyzed, so as to obtain the threshold of the Y-axis from the second image to be analyzed.

Specifically, one of the plurality of analyzed images corresponding to the second parameter can be randomly selected as the first image to be analyzed, as shown in Figure 3, which is the randomly selected first image 41 of FoxP3 to be analyzed. , all cells with the same fluorescent dye type (such as FoxP3) and in the same ROI are displayed, and each cell can be mapped to the analysis data. Multiple first click coordinates 411 can be clicked in the first image 41 to be analyzed (Figure 3 takes 5 clicks as an example), and each first click coordinate 411 can be mapped back to "cells in the ROI" in the analysis data. "XY coordinate" field, thereby knowing the cell corresponding to each first clicked coordinate 411. In this way, the signal strength corresponding to each first clicked coordinate 411 can be known, and the final Weak signal strength is used as the threshold on the x-axis.

Similarly, one of the plural analysis images corresponding to the third parameter can be randomly selected as the second image to be analyzed, as shown in Figure 4, which is the randomly selected second image to be analyzed 42 of Granzyme B. , all cells with the same fluorescent dye type (such as Granzyme B) and in the same ROI are displayed, and each cell can be mapped to the analysis data. Multiple second click coordinates 421 can be clicked in the second image 42 to be analyzed (Figure 4 takes five clicks as an example), and each second click coordinate 421 can be mapped back to "cells in the ROI" in the analysis data. "XY coordinate" field, thereby knowing the cell corresponding to each second clicked coordinate 421. In this way, the signal strength corresponding to each second clicked coordinate 421 can be known, and the final Weak signal strength is used as the Y-axis threshold.

In the above embodiment, the first/second click coordinate is the "XY coordinate of the cell in the ROI" as an example for explanation, but the present invention is not limited to this. If the first/second clicked coordinates cannot correspond to the "XY coordinates of the cells in the ROI" (for example, the coordinates read directly by clicking the mouse cursor do not exactly match the coordinates in the analysis data), you can click the first/second point Select the positive and negative specific values of the coordinates (such as x, y Search whether there is a cell within the interval of ±5) of the scalar value, thereby mapping the first/second click coordinates to the searched cell. Then proceed to step S4.

In step S4, the values corresponding to the second parameter and the third parameter are plotted on the graph to present a plurality of mark points corresponding to the first parameter in the graph.

Then, the plurality of marker points are distributed in the plurality of quadrants according to the threshold of the X-axis and the Y-axis. As shown in Figure 5, the X-axis threshold 21 can be about 0.04, and the Y-axis threshold 22 can be about 0.05. In this way, four quadrants can be defined: the first quadrant 31, the second quadrant 32, the third quadrant 33 and the Four Quadrants34. In addition, if one of the X-axis threshold and the Y-axis threshold is 0, then two quadrants can be defined, and the present invention does not limit the number of quadrants.

In other embodiments, the threshold of the X-axis and the threshold of the Y-axis can also be defined in other ways. For example, the threshold of the X-axis is the average of all values corresponding to the second parameter, the median of all values, or Is a custom value between 0 and 1. The threshold of the Y-axis is the average of all values corresponding to the third parameter, the position among all values, or a custom value between 0 and 1. , the custom value is a threshold value determined by the user after observing the distribution of multiple marker points. In this case, the present invention can omit the above step S3.

In order to make it easier for the user to conduct subsequent experiments, after completing step S4, that is, after distributing the plurality of marker points in multiple quadrants, step S5 can be performed to define different colors for the marker points in different quadrants. For example, the first quadrant 31 All marker points in the second quadrant 32 (FoxP3 ^{- /} Granzyme B ^{+ )} are defined as red, and all marker points in the third quadrant 33 (FoxP3 ^- /Granzyme B ^- ) is defined as black, and all marker points in the fourth quadrant 34 (FoxP3 ⁺ /Granzyme B ^- ) are defined as blue, etc., and based on the XY coordinates of the cells corresponding to the plurality of marker points in the tissue section image, Output the marker points of different colors into a cell distribution map corresponding to the tissue section image (as shown in Figure 6), or append multiple marker points and their corresponding colors to the analysis data (such as adding new fields) , to output an updated analysis data. In this way, users can further design multiple staining experiments based on the cell phenotype, number and proportion corresponding to the marker points in different quadrants, and can also define cells with Granzyme B ⁺ performance from CD8 ⁺ T cells. cells.

In one embodiment, the multiple fluorescent tissue image analysis method of the present invention can be executed on a computer using Anaconda software, or on a computer, server, cloud server, laptop, tablet or mobile phone Other types of software or programming languages are used in other hardware, and the present invention is not limited thereto.

Please refer to FIG. 2 . The multiple fluorescent tissue image processing system 1 of the present invention includes an input module 11 , a filtering and coordinate definition module 12 , a threshold definition module 13 , a graphics rendering module 14 and an output module 15 . The modules referred to in this article are specifically program code fragments, software or firmware for microprocessor execution. The functions of each module are the same as the technical content described in the aforementioned multiple fluorescent tissue image processing method, and will not be described again below.

The input module 11 is used to obtain analysis data of a tissue slice image with multiple cells and multiple analysis images corresponding to the analysis data. Each column in the analysis data represents the data of one cell, and each column can correspond to multiple columns. Fields may include: ROI number where each cell is located, tissue type (such as PanCK ⁺ or PanCK ^- ), cell phenotype (such as CD8 ⁺ T cell or CD4 ⁺ T cell, or Cytotoxic T cell defined by CD8 , Helper T cell defined by CD4, NK cell defined by CD56, etc.), cell ID, XY coordinates of the cell in the ROI, the nucleus, cytoplasm, cell membrane of each cell, multiple fluorescent dye types and autologous of the entire cell Fluorescent and so on. The analysis data can be, for example, a txt file or excel file output by the Phenoptics ^TM Tissue Analysis Software system software, but the present invention is not limited thereto. In addition, plural analysis images (pathology pathology maps) are classified by different fluorescent dye types and their ROIs. In other words, each analysis image contains all cells with the same fluorescent dye type and in the same ROI (for example, by analyzing Correspond to the ROI number in the data, the XY coordinates of the cells in the ROI, and the type of fluorescent dye).

The screening and coordinate definition module 12 is used to screen out cells with at least one first parameter from the analysis data, and to separate the second parameter and the third parameter in the cells with the first parameter after screening. As the X-axis and Y-axis of a graph, the first parameter is cell phenotype, and the cell phenotype can be CD8 ⁺ T cell or CD4 ⁺ T cell, and the second parameter or the third parameter is cell nucleus, cytoplasm, Multiple fluorescent dye types or autofluorescence of cell membranes, entire cells, but the invention is not limited thereto. In one embodiment, FoxP3 in the nucleus can be selected as the X-axis, and Granzyme B in the cytoplasm can be selected as the Y-axis, to further obtain the column of FoxP3 in the nucleus and Granzyme B in the cytoplasm of each cell obtained after screening. The value corresponding to the bit. In addition, after using the second parameter and the third parameter as the X-axis and Y-axis of the graph respectively, the filtering and coordinate definition module 12 distinguishes the filtered plurality of cells with the first parameter using a fourth parameter, where , the fourth parameter is the tissue type, which is PanCK ⁺ or PanCK ^- . In addition, while using the second parameter and the third parameter in the filtered plurality of cells with the first parameter as the X-axis and Y-axis of the graph respectively, the screening and coordinate definition module 12 can also use the second parameter and the third parameter as the X-axis and Y-axis of the graph. The values corresponding to the parameters are standardized and all values are compressed between 0 and 1 so that the fluorescence intensity of each cell marker has a consistent basis.

The threshold definition module 13 is used to select one of the plurality of analyzed images corresponding to the second parameter as the first image to be analyzed, to obtain the threshold value of the X-axis from the first image to be analyzed, and from the corresponding Select one of the plurality of analyzed images with the third parameter as the second desired Analyze the image to obtain the threshold value of the Y-axis from the second image to be analyzed. The detailed steps for obtaining the threshold values of the X-axis and Y-axis are described below.

Specifically, the threshold definition module 13 randomly selects one of the plurality of analyzed images corresponding to the second parameter as the first image to be analyzed. The selected first image 41 to be analyzed can be as shown in Figure 3. Click multiple first click coordinates 411 in the first image 41 to be analyzed (Figure 3 takes five clicks as an example). The threshold definition module 13 can map each first click coordinate 411 back to the analysis data. The "XY coordinates of the cell in the ROI" field is used to know the cell corresponding to each first click coordinate 411. In this way, the threshold definition module 13 can know each first click coordinate. 411 corresponding signal strength, and use the weakest signal strength as the threshold of the X-axis.

Similarly, the threshold definition module 13 randomly selects one of the plurality of analyzed images corresponding to the third parameter as the second image to be analyzed. The selected second image 42 to be analyzed can be as shown in Figure 4, and can be found in If multiple second click coordinates 421 are clicked in the second image 42 to be analyzed (Figure 4 takes five clicks as an example), the threshold definition module 13 can map each second click coordinate 421 back to the analysis data. The XY coordinates of the cell in the ROI" field is used to know the cell corresponding to each second click coordinate 421. In this way, the threshold definition module 13 can know each second click coordinate 411. The corresponding signal strength, and the weakest signal strength is used as the threshold of the Y-axis.

In the above embodiment, the first/second selected coordinates are the XY coordinates of the cells in the ROI as an example for explanation, but the present invention is not limited to this. If the first/second click coordinates cannot correspond to the XY coordinates of the cell in the ROI, the threshold definition module 13 can be in the range of the positive and negative specific values of the first/second click coordinates (for example, ±5 of the x, y coordinate values) Search for cells within the cell, thereby mapping the first/second click coordinates to the searched cells.

In other embodiments, the threshold definition module 13 can further define the thresholds of the X-axis and the Y-axis in the following method: the X-axis threshold is the average of all values corresponding to the second parameter, the median of all values, or customized Value, the Y-axis threshold is the average of all values corresponding to the third parameter, the median of all values, or a custom value. The present invention is not limited to this.

The graph drawing module 14 is used to draw the values corresponding to the second parameter and the third parameter in the graph, so as to present a plurality of mark points corresponding to the first parameter in the graph, and according to the threshold value of the X-axis and The threshold value of the Y-axis distributes the plurality of marker points in a plurality of quadrants, where the value corresponding to the second parameter or the third parameter is the signal intensity of multiple fluorescent dyes in the nucleus, cytoplasm, cell membrane, the entire cell, or autologous Fluorescence signal intensity, and as shown in Figure 5, the plural quadrants can be four quadrants, but they can also be two quadrants.

The output module 15 is used to define different colors for the marker points in different quadrants, and output the marker points of different colors into a corresponding tissue slice image based on the XY coordinates of the cells corresponding to the plurality of marker points in the tissue slice image. Cell distribution chart (as shown in Figure 6), or appending the plurality of marker points and their corresponding colors to the analysis data to output updated analysis data (for example, adding a new column).

In summary, the multiple fluorescent tissue image processing method and system of the present invention can easily screen out the cells to be analyzed. For example, the fluorescent signal generated by Opal dye will be displayed on the cell membrane, cytoplasm, nucleus and the entire cell. The present invention These fluorescence data can be further analyzed. According to various cell markers and the position on the cell stain (cell membrane, cytoplasm or nucleus), cells with well-defined signals at reasonable positions can be selected, and then more definitions can be given. It is convenient for users to do subsequent data processing. In addition, the imaging cytometry defined in the present invention can output more intuitive analysis results, such as the proportion of daughter cell groups in main cells and the distribution map of daughter cell groups in ROI images, which can facilitate users to compare and return existing fluorescent data. Use the pathological diagram provided by the optical image analysis tool to determine whether its definition is correct. In addition, the present invention can also establish a threshold for each fluorescent signal according to the signal intensity, and define cells above the threshold as positive and cells below the threshold as negative, thereby analyzing the multiple fluorescent positivity of each cell in the graph. Signals can be used to distinguish the definitions of multiple fluorescent signals on a single cell. There is no need to spend time looking for >= 5 cells to define, achieving the purpose of defining multiple cell markers for a single cell, solving the problem that in the previous technology, only a single cell can be defined A problem of fluorescence signal definition is provided to provide users with a better understanding of the phenotypic status of each cell, and facilitate the analysis of daughter cell types of a certain cell group, and can obtain the results in a shorter time than existing fluorescence image analysis tools. Multiple fluorescence definitions of a single cell can provide results that are more in line with individual needs.

The above embodiments are only for illustrating the technical principles, characteristics and effects of the present invention, and are not intended to limit the scope of the present invention. Anyone familiar with this technology can implement the present invention without violating the spirit and scope of the present invention. Modifications and changes are made to the above embodiments. However, any equivalent modifications and changes accomplished by applying the teachings of the present invention should still be covered by the following patent application scope. The scope of protection of the rights of the present invention shall be as listed in the following patent application scope.

S1-S5: Steps

Claims (13)

A multiple fluorescence tissue image analysis method, including: obtaining analysis data of a tissue slice image with a plurality of cells and a plurality of analysis images corresponding to the analysis data; screening out those cells with at least one first parameter from the analysis data , and use the second parameter and the third parameter in the plurality of cells with the first parameter after screening as the X-axis and Y-axis of a graph respectively; select one of the plurality of analysis images corresponding to the second parameter. One of them is used as the first image to be analyzed, and a plurality of first click coordinates are obtained from the first image to be analyzed, and the weakest signal intensity among the cells corresponding to the plurality of first click coordinates is used as the threshold of the X-axis , and select one of the plurality of analysis images corresponding to the third parameter as the second image to be analyzed, and obtain the plurality of second point selection coordinates from the second image to be analyzed, so as to combine the plurality of second points Select the weakest signal strength among the cells corresponding to the coordinates as the threshold of the Y-axis; and draw the values corresponding to the second parameter and the third parameter in the graph to present the corresponding first parameter in the graph A plurality of marked points, and the plurality of marked points are distributed in a plurality of quadrants according to the threshold of the X-axis and the threshold of the Y-axis. The multiple fluorescent tissue image analysis method as described in claim 1, wherein the first parameter is a cell phenotype, and the cell phenotype is CD8 ⁺ T cell or CD4 ⁺ T cell. The multiple fluorescent tissue image analysis method as described in claim 1 further includes, after using the second parameter and the third parameter as the X-axis and Y-axis of the graph respectively, filtering the first parameter. The plurality of cells are further distinguished by a fourth parameter, wherein the fourth parameter is a tissue type, and the tissue type is PanCK ⁺ or PanCK ⁻ . The multiple fluorescent tissue image analysis method as described in claim 1, wherein the second parameter or the third parameter is a plurality of fluorescent dye types or autofluorescence of the cell nucleus, cytoplasm, cell membrane, the entire cell, and the The value corresponding to the second parameter or the third parameter is the signal intensity of multiple fluorescent dyes or the signal intensity of autofluorescence in the cell nucleus, cytoplasm, cell membrane, or the entire cell. The multiple fluorescent tissue image analysis method as described in claim 1, further comprising: using the second parameter and the third parameter in the plurality of cells having the first parameter after screening as the X-axis of the graph respectively. and the Y-axis, normalize the values corresponding to the second parameter and the third parameter. The multiple fluorescence tissue image analysis method as described in claim 1 further includes: defining different colors for the marker points in different quadrants, and based on the XY coordinates of the cells corresponding to the plurality of marker points in the tissue slice image, The marked points of different colors are output into a cell distribution map corresponding to the tissue section image. The multiple fluorescence tissue image analysis method as described in claim 6 further includes: appending the plurality of marker points and their corresponding colors to the analysis data to output updated analysis data. A multiple fluorescent tissue image processing system, including: an input module for obtaining analysis data of a tissue slice image with a plurality of cells and a plurality of analysis images corresponding to the analysis data; a screening and coordinate definition module for obtaining the analysis data from the Analyze the data to screen out those cells that have at least one first parameter, and use the second parameter and the third parameter in the screened cells that have the first parameter as the X-axis and Y-axis of a graph respectively. ; The threshold definition module is used to select one of the plurality of analysis images corresponding to the second parameter as the first image to be analyzed, and to obtain the plurality of first click coordinates from the first image to be analyzed, to will The weakest signal intensity among the cells corresponding to the plurality of first click coordinates is used as the threshold of the X-axis, and one of the plurality of analysis images corresponding to the third parameter is selected as the second image to be analyzed, from A plurality of second click coordinates are obtained from the second image to be analyzed, and the weakest signal intensity among the cells corresponding to the plurality of second click coordinates is used as the threshold of the Y-axis; and a graphics drawing module is used to convert the The values corresponding to the second parameter and the third parameter are plotted in the graph to present a plurality of mark points corresponding to the first parameter in the graph, and the plurality of points are divided according to the threshold value of the X-axis and the threshold value of the Y-axis. Marked points are distributed in complex quadrants. The multiple fluorescent tissue image processing system of claim 8, wherein the first parameter is a cell phenotype, and the cell phenotype is CD8 ⁺ T cell or CD4 ⁺ T cell. The multiple fluorescent tissue image processing system as described in claim 8, wherein the filtering and coordinate definition module will be filtered after using the second parameter and the third parameter as the X-axis and Y-axis of the graph respectively. The plurality of cells having the first parameter are then distinguished by a fourth parameter, wherein the fourth parameter is a tissue type, and the tissue type is PanCK ⁺ or PanCK ⁻ . The multiple fluorescent tissue image processing system as described in claim 8, wherein the second parameter or the third parameter is a plurality of fluorescent dye types or autofluorescence of the cell nucleus, cytoplasm, cell membrane, the entire cell, and the The value corresponding to the second parameter or the third parameter is the signal intensity of multiple fluorescent dyes or the signal intensity of autofluorescence in the cell nucleus, cytoplasm, cell membrane, or the entire cell. The multiple fluorescent tissue image processing system as described in claim 8, wherein the screening and coordinate definition module uses the second parameter and the third parameter in the plurality of cells having the first parameter after screening, respectively. As the X-axis and Y-axis of the graph, the values corresponding to the second parameter and the third parameter are normalized. The multiple fluorescent tissue image processing system as described in claim 8 further includes an output module for defining different colors for the marker points in different quadrants, and according to the cells corresponding to the plurality of marker points in the tissue slice image. XY coordinates, the marker points of different colors are output into a cell distribution map corresponding to the tissue section image, or the plurality of marker points and their corresponding colors are appended to the analysis data to output an updated analysis data.

Images