A high-throughput single-cell image statistical analysis method and system

Abstract

The application discloses a high-throughput single-cell image statistical analysis method and system, and the method comprises the following steps: after cells are introduced into a microfluidic chip, the cells are fluorescently labeled, and an image of the microfluidic chip is acquired after the fluorescent labeling; image preprocessing is performed on the acquired small groove image and flow channel region image; whether single cells exist in the small grooves is identified based on the preprocessed small groove image through an image recognition model; when it is identified that single cells exist in the small grooves, each preprocessed flow channel region image corresponding to each small groove where single cells exist is intercepted; each flow channel region image is analyzed and processed to calculate the brightness value of each flow channel region image; and based on the brightness value of each flow channel region image and each small groove where single cells exist, position, quantity and fluorescence signal intensity information of the required single cells are statistically analyzed. The application can screen all required cells from tens of thousands of cells.

Description

Technical Field

[0001] This invention belongs to the field of single-cell detection technology, specifically relating to a high-throughput single-cell image statistical analysis method and system. Background Technology

[0002] In biomedical research and drug development, rapid, large-scale, and accurate screening and identification of specific cell functions and characteristics are of great significance. Existing cell screening methods mainly rely on manual observation and judgment, which is inefficient and easily influenced by subjective factors. Therefore, developing an automated, high-precision, and high-throughput cell screening method is a pressing technical problem that needs to be solved. Summary of the Invention

[0003] The purpose of this invention is to provide a high-throughput single-cell image statistical analysis method and system to solve the problems or deficiencies existing in the prior art.

[0004] To achieve the above objectives, the present invention employs the following technical solution:

[0005] This invention provides a high-throughput single-cell image statistical analysis method, comprising:

[0006] After cells are introduced into a microfluidic chip, they are fluorescently labeled, and images of the microfluidic chip are acquired after fluorescent labeling. The microfluidic chip includes several small grooves and flow channel regions corresponding to the small grooves. The acquired images of the microfluidic chip include images of the small grooves and images of the flow channel regions. The acquired images of the microfluidic chip contain fluorescently labeled cells.

[0007] The acquired images of the small groove and the flow channel region are preprocessed, including at least image enhancement, noise reduction and segmentation.

[0008] Based on the preprocessed groove image, an image recognition model is used to identify whether there are single cells in the groove.

[0009] Once a small groove containing a single cell is identified, the image of each flow channel region above each groove containing a single cell, after preprocessing, is extracted.

[0010] The image of each flow channel region is analyzed and processed to calculate the brightness value of each flow channel region image.

[0011] Based on the brightness value of the image of each flow channel region and each small groove containing a single cell, the location, number, and fluorescence signal intensity information of the required single cells are statistically analyzed.

[0012] Optionally, the cell situation in the groove can be identified by an image recognition model, wherein the cell situation in the groove includes no cells in the groove, a single cell in the groove, and / or multiple cells in the groove.

[0013] Optionally, the image of each flow channel region is analyzed and processed to calculate the brightness value of each flow channel region image, including:

[0014] Convert the image of each flow channel region to HSV format;

[0015] After format conversion, the brightness value of each flow channel region image is calculated. This brightness value is then compared with a preset brightness value to filter out the desired single cells.

[0016] The brightness values ​​of the images of each flow channel region are sorted, and the required single cells are identified and selected according to the preset brightness sorting rules.

[0017] Optionally, image preprocessing is performed on the acquired slot images and flow channel region images, including:

[0018] Light compensation is applied to the images of the small grooves and the flow channel region to achieve a uniform light effect;

[0019] After performing light compensation, the fluorescent background in the slot image and the flow channel area image is removed;

[0020] After removing the fluorescent background from the slot image and the flow channel image, a noise reduction algorithm is used to reduce the random noise in the slot image and the flow channel image.

[0021] Optionally, after fluorescently labeling the cell secondary antibody in the microfluidic chip, the microfluidic chip is divided into several regions according to the field of view, and images of all regions of the microfluidic chip are captured by a CCD camera. The microfluidic chip is rectangular.

[0022] Optionally, the motion platform can be controlled to move in a straight line via control software, so that the microfluidic chip located on the motion platform can move in a straight line. At the same time, the CCD camera can be controlled to take a panoramic picture of the microfluidic chip while the microfluidic chip is moving in a straight line, so as to obtain an image of the microfluidic chip. The CCD camera is located above the motion platform and the lens is facing the microfluidic chip.

[0023] Optionally, after the desired single cells are selected, the number of the slot where the single cell is located is recorded to obtain the specific location of the desired single cell.

[0024] Optionally, cells in the flow channel region can be pulled into the small groove using photoelectric tweezers.

[0025] Accordingly, this application also provides a high-throughput single-cell image statistical analysis system, including:

[0026] The image acquisition module is used to fluorescently label the cells after they are introduced into the microfluidic chip, and to acquire an image of the microfluidic chip after fluorescent labeling. The microfluidic chip includes several small grooves and flow channel regions corresponding to the small grooves. The acquired image of the microfluidic chip includes images of the small grooves and images of the flow channel regions. The acquired image of the microfluidic chip contains fluorescently labeled cells.

[0027] The preprocessing module is used to perform image preprocessing on the acquired slot images and flow channel region images. The preprocessing includes at least image enhancement, noise reduction, and segmentation.

[0028] The recognition module is used to identify whether there are single cells in the groove based on the preprocessed groove image using an image recognition model.

[0029] The flow channel region image screenshot module is used to extract the preprocessed flow channel region image above each small groove containing a single cell after identifying the small groove containing a single cell.

[0030] The calculation module is used to analyze and process the image of each flow channel region to calculate the brightness value of each flow channel region image.

[0031] The statistical analysis module is used to statistically analyze the location, number, and fluorescence signal intensity information of the required single cells based on the brightness value of the image of each flow channel region and each small groove where single cells exist.

[0032] Optionally, the calculation module is specifically used for:

[0033] Convert the image of each flow channel region to HSV format;

[0034] After format conversion, the brightness value of each flow channel region image is calculated. This brightness value is then compared with a preset brightness value to filter out the desired single cells.

[0035] The brightness values ​​of the images of each flow channel region are sorted, and the required single cells are identified and selected according to the preset brightness sorting rules.

[0036] Beneficial effects:

[0037] 1. This application's screening method can solve the problem of low efficiency in manually viewing fluorescence results and counting cells. It has a high degree of automation, reduces manual intervention, and improves screening efficiency.

[0038] 2. High-throughput cell screening can simultaneously detect and identify a large number of cells, and the entire process can screen out all the required cells from tens of thousands of cells.

[0039] 3. High precision: Based on image processing algorithms, it accurately identifies cells that secrete the corresponding antibodies.

[0040] 4. Enhance data analysis capabilities, provide rich visualization results and data analysis tools, which help to deeply explore experimental data and assist scientific research decision-making. Attached Figure Description

[0041] The accompanying drawings, which form part of this application, are used to provide a further understanding of the application and to make other features, objects, and advantages of the application more apparent. The illustrative embodiments and descriptions of this application are used to explain the application and do not constitute an undue limitation of the application. In the drawings:

[0042] Figure 1 A flowchart illustrating a high-throughput single-cell image statistical analysis method provided in this application embodiment;

[0043] Figure 2 This is a partial structural schematic diagram of a microfluidic chip provided in an embodiment of this application;

[0044] Figure 3 This is a schematic diagram of the structure of a high-throughput single-cell image statistical analysis system provided in an embodiment of this application. Detailed Implementation

[0045] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the present invention will be briefly introduced below in conjunction with the accompanying drawings and descriptions of the embodiments or the prior art. Obviously, the following description of the structure of the accompanying drawings is only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. It should be noted that the description of these embodiments is for the purpose of helping to understand the present invention, but does not constitute a limitation of the present invention.

[0046] It should be understood that although the terms first, second, etc., may be used herein to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit, without departing from the scope of the exemplary embodiments of the invention.

[0047] It should be understood that the term "and / or" that may appear in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, B exists alone, and A and B exist simultaneously. The term " / and" that may appear in this document describes another relationship between related objects, indicating that two relationships can exist. For example, A / and B can mean: A exists alone, and A and B exist alone. In addition, the character " / " that may appear in this document generally indicates that the related objects before and after it are in an "or" relationship.

[0048] Example 1:

[0049] like Figure 1 The diagram shown is a flowchart of a high-throughput single-cell image statistical analysis method proposed in an embodiment of the present invention, including:

[0050] S1. After introducing cells into the microfluidic chip, the cells are fluorescently labeled, and an image of the microfluidic chip is acquired after fluorescent labeling.

[0051] Specifically, to screen for the desired cells, the cells are introduced into a microfluidic device (i.e., a microfluidic chip), then fluorescently labeled with a secondary antibody. Images of the microfluidic chip are then acquired after the fluorescent labeling process, for reference. Figure 2 Microfluidic devices contain a large number of small grooves. Figure 2 The image shown is only a partial diagram of the microfluidic device (the entire groove structure is not shown). The microfluidic chip includes several grooves and corresponding flow channel regions. After cells are introduced into the microfluidic device, many cells will initially be present in the lateral region (flow channel). Using photoelectric tweezers, individual cells are dragged into the grooves below. It is possible to drag in one cell, multiple cells, or none at all. The acquired images of the microfluidic chip include images of the grooves and flow channel regions. Fluorescently labeled cells are present in the acquired images of the microfluidic chip.

[0052] In one embodiment of this application, the movement of a microfluidic chip on a motion platform is controlled by control software, as well as the capture of images by a CCD camera. The microfluidic chip is placed on the motion platform, and a CCD camera is positioned above the motion platform. Since the CCD camera cannot capture the entire area of ​​the chip at once, the chip is moved by the motion platform so that the CCD camera can capture the entire area of ​​the microfluidic chip.

[0053] Fluorescent secondary antibodies: Primary antibodies are antibodies specifically produced against cells. Secondary antibodies are antibodies that use the primary antibody as an antigen. During cell sorting, the binding of the secondary antibody to the primary antibody indirectly labels the target antigen, allowing visualization under a fluorescence microscope. Fluorescent secondary antibodies are an important method in cell biology research. Cell surface antigens or other molecules can be labeled with fluorescently labeled antibodies. These antibodies bind to specific cell surface antigens, and cells can be sorted based on different fluorescence intensities or colors to obtain a subset of cells expressing specific markers.

[0054] S2, perform image preprocessing on the acquired slot image and flow channel region image, the preprocessing including at least image enhancement, noise reduction and segmentation.

[0055] To accurately preprocess the acquired images, in a preferred embodiment of this solution, image preprocessing is performed on the acquired slot images and flow channel region images, including:

[0056] Light compensation is applied to the images of the small grooves and the flow channel region to achieve a uniform light effect;

[0057] After performing light compensation, the fluorescent background in the slot image and the flow channel area image is removed;

[0058] After removing the fluorescent background from the slot image and the flow channel image, a noise reduction algorithm is used to reduce the random noise in the slot image and the flow channel image.

[0059] Specifically, image preprocessing is performed. The preprocessing methods are: (1) Light compensation: Due to the characteristics of microscopic images, there will be a situation where the center is bright and the surrounding area is dark. A mask is superimposed to perform light compensation (the mask image has been calculated based on the image after the optical path is built and before the fluorescence experiment is performed). (2) Background correction: Remove the background fluorescence in the image. This is achieved by calculating the local or global background average value of the image and then subtracting it from the original image. (3) Noise filtering: Apply the noise reduction algorithm Gaussian filtering to reduce random noise in the image.

[0060] S3, based on the preprocessed small groove image, uses an image recognition model to identify whether there are single cells in the small groove.

[0061] Specifically, after preprocessing the slot images, an image recognition model is used to identify whether single cells exist within the slots. The construction process of the image recognition model is as follows: Training sets, datasets, and validation sets are obtained from the acquired images. The acquired images are labeled with single-cell images, multi-cell images, and images with impurities. The labeled training set is used for training, the test set is used to test the model, and the validation set is used for validation. After successful validation, the final image recognition model is obtained. The trained image recognition model can directly determine which locations (pixel positions) within the slots contain cells, and whether they are single or multi-celled. The pixel positions of the slots are known. The image recognition model identifies the presence of cells within the slots, including the absence of cells, the presence of single cells, and / or the presence of multi-celled cells. After selecting the desired single cell, the slot number information is recorded to obtain the specific location of the desired single cell.

[0062] S4. After identifying the small groove containing a single cell, extract the image of each preprocessed flow channel region above each small groove containing a single cell.

[0063] Specifically, after identifying the slots containing single cells using the image recognition model, the preprocessed image of each flow channel region above each slot containing a single cell is extracted. The fluorescence intensity is calculated; cells with higher intensity are the desired cells, and their slot numbers are recorded.

[0064] S5 analyzes and processes the image of each flow channel region to calculate the brightness value of each flow channel region image.

[0065] To accurately obtain the brightness value of each flow channel region image, in a preferred embodiment of this solution, the image of each flow channel region is analyzed and processed to calculate the brightness value of each flow channel region image, including:

[0066] Convert the image of each flow channel region to HSV format;

[0067] After format conversion, the brightness value of each flow channel region image is calculated. This brightness value is then compared with a preset brightness value to filter out the desired single cells.

[0068] The brightness values ​​of the images of each flow channel region are sorted, and the required single cells are identified and selected according to the preset brightness sorting rules.

[0069] Specifically, after converting the image to HSV format, image processing algorithms are used to obtain the brightness value (i.e., fluorescence signal intensity). HSV format is a color representation model that describes color based on three basic attributes: hue, saturation, and value. The value v represents brightness, and the mean v is calculated to represent the overall brightness of the image. After format conversion, the brightness value of each flow channel region is calculated. This brightness value is compared with a preset brightness value to select the desired single cells. Alternatively, the brightness values ​​of each flow channel region are sorted, and the desired single cells are identified and selected according to a preset brightness sorting rule.

[0070] The brightness value is compared with a preset brightness value to filter out the required single cells. For example, if the brightness value of the flow channel region image corresponding to a certain slot (where a single cell exists) is higher than the preset brightness threshold, then the single cell in the slot is determined as the required single cell.

[0071] According to the preset brightness sorting rules, the required single cells are identified and screened out. For example, multiple small grooves (which contain single cells) are sorted in order from high to low brightness values ​​of the flow channel area image to obtain a small groove sequence. Then, the first K small grooves in the small groove sequence are selected, and finally, the single cells in the first K small grooves are taken as K required single cells, where K represents a positive integer.

[0072] S6. Based on the brightness value of each flow channel region image and each small groove containing a single cell, statistically analyze the location (groove number), quantity (number of grooves), and fluorescence signal intensity information (brightness value of the flow channel region image corresponding to the groove) of the required single cell.

[0073] Specifically, since the location of the slots is known, the image recognition model can determine which cell each slot contains, thus identifying its corresponding brightness value. The slot number is recorded, and all data, including the location, number, and fluorescence signal intensity of the desired single cell, is compiled. This application can accurately determine the quantity and / or quality (the degree to which a set of inherent properties of the object satisfies requirements) of the analyte generated by micro-objects confined within slots in a microfluidic device.

[0074] Example 2:

[0075] To further illustrate the technical concept of this invention, the technical solution of this invention will be described in conjunction with specific application scenarios.

[0076] 1. The software reads the image data of the microfluidic chip that has been stored on the computer.

[0077] 2. Image preprocessing is performed. The preprocessing methods are: (1) Light compensation: Due to the characteristics of microscopic images, there will be a situation where the center is bright and the surrounding area is dark. A mask is superimposed to perform light compensation (the mask image has been calculated based on the image after the optical path is built and before the fluorescence experiment is performed). (2) Background correction: Remove the background fluorescence in the image. This is achieved by calculating the local or global background average value of the image and then subtracting it from the original image. (3) Noise filtering: Apply the noise reduction algorithm Gaussian filtering to reduce random noise in the image.

[0078] 3. Use the trained image recognition model to identify the preprocessed slotted image. The trained image recognition model can directly determine which locations (pixel positions) within the slot contain cells, and whether they are single cells or multiple cells. After selecting the desired single cell, record the slot number information where the single cell is located to obtain the specific location of the desired single cell.

[0079] 4. Capture the image at the corresponding location (the flow channel area above the small groove containing the single cell), convert the captured image to HSV format, and calculate the mean value of its V as its brightness. Convert the image to HSV format, where the v value represents brightness. Calculate the mean value of v; this mean value represents the brightness of the entire image. Sort the brightness values ​​of all images. You can choose to sort only the top few values, or set a desired range for brightness value differences; this is customizable.

[0080] The image cropping is implemented internally by the software and does not require manual operation. The entire image contains many small grooves, and the position of each groove in the image is recorded. The flow channel image above the groove at the corresponding pixel position can be directly cropped.

[0081] 5. With the location of the slots known, the image recognition model will identify the pixel location of the cells. Based on this, it is possible to determine which cell each slot contains, know its corresponding brightness value, record the slot number, and compile all the data.

[0082] This application solves the problem of low efficiency in manually viewing fluorescence results and counting cells. The screening method of this application is highly automated, reducing manual intervention and improving screening efficiency. It can simultaneously detect and identify a large number of cells, and the entire process can screen out all the desired cells from tens of thousands of cells.

[0083] Figure 3 A schematic diagram of the structure of a high-throughput single-cell image statistical analysis system provided by the present invention includes:

[0084] The image acquisition module 10 is used to fluorescently label the cells after they are introduced into the microfluidic chip, and to acquire an image of the microfluidic chip after fluorescent labeling. The microfluidic chip includes several small grooves and flow channel regions corresponding to the small grooves. The acquired image of the microfluidic chip includes images of the small grooves and images of the flow channel regions. The acquired image of the microfluidic chip contains fluorescently labeled cells.

[0085] The preprocessing module 20 is used to perform image preprocessing on the acquired slot image and flow channel region image, the preprocessing including at least image enhancement, noise reduction and segmentation;

[0086] The recognition module 30 is used to identify whether there are single cells in the groove based on the preprocessed groove image using an image recognition model;

[0087] The flow channel region image screenshot module 40 is used to extract the preprocessed flow channel region image above each small groove containing a single cell after identifying the small groove containing a single cell.

[0088] The calculation module 50 is used to analyze and process the image of each flow channel region to calculate the brightness value of the image of each flow channel region.

[0089] The statistical analysis module 60 is used to statistically analyze the location, number, and fluorescence signal intensity information of the required single cells based on the brightness value of the image of each flow channel region and each small groove where a single cell exists.

[0090] The calculation module is specifically used to: convert the image of each flow channel region into HSV format; calculate the brightness value of each flow channel region image after the format conversion is completed, compare the brightness value with a preset brightness value to filter out the required single cells, or sort the brightness values ​​of each flow channel region image, and identify and filter out the required single cells according to a preset brightness sorting rule.

[0091] The high-throughput single-cell image statistical analysis system of this application can enhance data analysis capabilities, provide rich visualization results and data analysis tools, help to deeply explore experimental data, and assist scientific research decision-making.

[0092] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

[0093] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A high-throughput single-cell image statistical analysis method, characterized in that, include: After cells are introduced into a microfluidic chip, they are fluorescently labeled, and images of the microfluidic chip are acquired after fluorescent labeling. The microfluidic chip includes several small grooves and flow channel regions corresponding to the small grooves. The acquired images of the microfluidic chip include images of the small grooves and images of the flow channel regions. The acquired images of the microfluidic chip contain fluorescently labeled cells. The acquired slot images and flow channel region images are subjected to image preprocessing, which includes at least image enhancement, noise reduction, and segmentation, including: Light compensation is applied to the images of the small grooves and the flow channel region to achieve a uniform light effect; After performing light compensation, the fluorescent background in the slot image and the flow channel area image is removed; After removing the fluorescent background from the slot image and the flow channel region image, a noise reduction algorithm is used to reduce the random noise in the slot image and the flow channel region image. Based on the preprocessed groove image, the presence of single cells in the groove is identified by an image recognition model. The cell situation in the groove is also identified by an image recognition model, including no cells in the groove, single cells in the groove, and / or multiple cells in the groove. Once a small groove containing a single cell is identified, the image of each flow channel region above each groove containing a single cell, after preprocessing, is extracted. The image of each flow channel region is analyzed and processed to calculate the brightness value of each flow channel region image, including: Convert the image of each flow channel region to HSV format; After format conversion, the brightness value of each flow channel region image is calculated. This brightness value is then compared with a preset brightness value to filter out the desired single cells. The brightness values ​​of the images of each flow channel region are sorted, and the required single cells are identified and selected according to the preset brightness sorting rules. Based on the brightness value of the image of each flow channel region and each small groove containing a single cell, the location, number, and fluorescence signal intensity information of the required single cells are statistically analyzed. After fluorescently labeling the secondary antibody for cells in the microfluidic chip, the microfluidic chip is divided into several regions according to the field of view. Images of all regions of the microfluidic chip are captured by a CCD camera. The microfluidic chip is rectangular. The control software controls the motion platform to move in a straight line, so that the microfluidic chip located on the motion platform can move in a straight line. At the same time, the CCD camera is controlled to take a panoramic picture of the microfluidic chip while it is moving in a straight line, so as to obtain an image of the microfluidic chip. The CCD camera is located above the motion platform and its lens is pointed towards the microfluidic chip.

2. The method according to claim 1, characterized in that, Once the desired single cells are selected, the cell number is recorded to obtain the specific location of the desired single cells.

3. The method according to claim 1, characterized in that, Cells in the flow channel region are pulled into the small groove using photoelectric tweezers.

4. A high-throughput single-cell image statistical analysis system, used to implement the method according to any one of claims 1 to 3, characterized in that, include: The image acquisition module is used to fluorescently label the cells after they are introduced into the microfluidic chip, and to acquire an image of the microfluidic chip after fluorescent labeling. The microfluidic chip includes several small grooves and flow channel regions corresponding to the small grooves. The acquired image of the microfluidic chip includes images of the small grooves and images of the flow channel regions. The acquired image of the microfluidic chip contains fluorescently labeled cells. The preprocessing module is used to perform image preprocessing on the acquired slot images and flow channel region images. The preprocessing includes at least image enhancement, noise reduction, and segmentation. The recognition module is used to identify whether there are single cells in the groove based on the preprocessed groove image using an image recognition model. The flow channel region image screenshot module is used to extract the preprocessed flow channel region image above each small groove containing a single cell after identifying the small groove containing a single cell. The calculation module is used to analyze and process the image of each flow channel region to calculate the brightness value of each flow channel region image. The statistical analysis module is used to statistically analyze the location, number, and fluorescence signal intensity information of the required single cells based on the brightness value of the image of each flow channel region and each small groove where single cells exist.

5. The system according to claim 4, characterized in that, The calculation module is specifically used for: Convert the image of each flow channel region to HSV format; After format conversion, the brightness value of each flow channel region image is calculated. This brightness value is then compared with a preset brightness value to filter out the desired single cells. The brightness values ​​of the images of each flow channel region are sorted, and the required single cells are identified and selected according to the preset brightness sorting rules.

Images