Calculation method and device of cell confluence degree and computer equipment

Abstract

The embodiment of the invention is applicable to the technical field of biomedicine and image processing, and provides a cell confluence degree calculation method, a cell confluence degree calculation device and computer equipment. According to the method, an adherent cell original image is processed by using opening operation, closing operation and the like in a morphological image processing algorithm, and contrast enhancement is combined, so that the confluence degree of cells is obtained. The original image is optimized by brightness equalization means such as gamma correction, the pixel gradient change of the stem cell region image can be enhanced, and the characteristics of the stem cell image are highlighted. And then, locking a cell region by using an edge detection algorithm, and obtaining a complete cell region through algorithms such as image connected domain calculation, flooding filling and optimized region growth. On the basis, stem cell image cell region segmentation and cell confluence degree calculation can be quickly realized based on a traditional algorithm. By applying the calculation method of the cell confluence degree provided by the embodiment of the invention, the calculation of the cell confluence degree can be quickly and accurately realized.

Description

Technical Field

[0001] The embodiments of the present application belong to the field of biomedicine and image processing technology, and in particular, relate to a method, device and computer equipment for calculating cell confluence. Background Art

[0002] Cell confluence refers to the ratio of the area occupied by cells to the surface area of ​​the culture after the cells grow adherently and fully expanded in the culture dish or flask. It is an important and necessary parameter to track and evaluate in primary cell culture.

[0003] Traditional cell confluence assessment methods are mainly carried out through manual observation. Judging cell confluence through manual observation requires high professionalism of personnel and is greatly influenced by the subjective influence of the assessors. Therefore, methods based on image processing technology to calculate cell confluence have emerged in the prior art. However, such technologies often use clustering algorithms or are implemented by building corresponding models. Not only is the calculation process more complex and inefficient, but it is also easily affected by data samples, resulting in a low accuracy of the final calculated confluence result. Summary of the invention

[0004] In view of this, the embodiments of the present application provide a method, apparatus and computer device for calculating cell confluence, so as to reduce the algorithm complexity of cell confluence calculation and improve the calculation speed and accuracy of the calculation results.

[0005] A first aspect of an embodiment of the present application provides a method for calculating cell confluence, comprising:

[0006] Collect adherent cell images of stem cells;

[0007] Performing optimization processing on the adherent cell image to obtain a target image, wherein the optimization processing includes enhancing the cell image features in the adherent cell image;

[0008] Determine the cell area in the target image by edge detection;

[0009] The cell confluence of the stem cells is calculated based on the cell area in the target image.

[0010] A second aspect of an embodiment of the present application provides a device for calculating cell confluence, comprising:

[0011] An image acquisition module, used for acquiring adherent cell images of stem cells;

[0012] An image optimization module, used for optimizing the adherent cell image to obtain a target image, wherein the optimization includes enhancing the cell image features in the adherent cell image;

[0013] An image segmentation module, used for determining the cell area in the target image by edge detection;

[0014] The cell confluence calculation module is used to calculate the cell confluence of the stem cells based on the cell region in the target image.

[0015] A third aspect of an embodiment of the present application provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for calculating cell confluence as described in the first aspect above is implemented.

[0016] A fourth aspect of an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for calculating cell confluence as described in the first aspect above is implemented.

[0017] A fifth aspect of the embodiments of the present application provides a computer program product. When the computer program product is run on a computer, the computer is enabled to execute the method for calculating cell confluence described in the first aspect.

[0018] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0019] The embodiment of the present application collects the adherent cell image of the stem cell, and can determine the cell area in the target image by edge detection on the basis of optimizing the adherent cell image to obtain the target image, so as to calculate the cell confluence of the stem cell based on the cell area in the target image. Compared with the artificial observation method used in the prior art to evaluate the cell confluence, the calculation method evaluation standard used in the embodiment of the present application is more objective and more capable of large-scale production operations. Secondly, the relevant algorithms and processing steps used in the embodiment of the present application are more capable of accurately detecting the complete cell area compared to the simple and fast threshold segmentation in the prior art; compared to the K-Means clustering algorithm with more complex logic, the embodiment of the present application needs to adjust fewer parameters during the application process, the algorithm effect is more stable, and the algorithm efficiency is also higher. Third, the application of the embodiment of the present application does not need to pre-make a data set, which can save a lot of manpower costs, and the calculation process also requires too much computing resources, and the requirements for equipment hardware conditions are relatively low. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments or prior art descriptions. Obviously, the drawings described below are only some embodiments of the present application, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0021] Figure 1 is a schematic diagram of a method for calculating cell confluence provided in an embodiment of the present application;

[0022] Figure 2 is a schematic diagram of optimizing the image of adherent cells provided in an embodiment of the present application;

[0023] Figure 3 is a schematic diagram of a possible implementation of S103 in a method for calculating cell confluence provided in an embodiment of the present application;

[0024] Figure 4 is a schematic diagram of segmenting a target image provided by an embodiment of the present application;

[0025] Figure 5 is a schematic diagram of a device for calculating cell confluence provided in an embodiment of the present application;

[0026] Figure 6 It is a schematic diagram of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0027] In the following description, specific details such as specific system structures, technologies, etc. are proposed for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from hindering the description of the present application.

[0028] As mentioned above, the cell confluence can be automatically calculated based on image processing technology. For example, the improved K-Means clustering algorithm can be used to segment the microscopic image of adherent cells, and then the cell confluence is calculated based on the segmented image. This type of algorithm is mainly used for the calculation of the confluence of umbilical cord mesenchymal stem cells. By optimizing the selection method of the initial clustering center, the accurate separation of the irregularly adherent umbilical cord mesenchymal stem cell microscopic image and the background can be achieved. When applying the K-Means clustering algorithm, it is necessary to iteratively calculate the distance between the sample point and the clustering center, and continuously adjust the position of the clustering center until the termination condition is met. Therefore, this type of algorithm has the defects of being sensitive to the initial clustering center, sensitive to noise and outliers, and difficult to select the K value. In addition, in the face of complex situations such as changes in illumination and poor imaging quality, the above algorithm does not have sufficient robustness and cannot cover most application scenarios.

[0029] Another method for calculating cell confluence is to construct a background model for high-definition cell images and use the background model to suppress the background of high-definition cell images, so as to achieve the purpose of data enhancement and be used for cell confluence calculation. This method can identify image corners based on the background suppressed image, build an optical flow model by combining the corners and continuous frame images, and use the constructed optical flow model to extract the temporal features of the suppressed image. By fusing the temporal features with the spatial features, the fused features can be used to analyze the cell confluence of mesenchymal stem cells. However, the above algorithm is highly complex and time-consuming. Building an optical flow model also requires a large amount of calculation, and has high requirements for device hardware.

[0030] In addition, a large number of sample images can be collected to divide them into training sets, test sets, and validation sets. On this basis, a cell recognition model is constructed, and the cell area in the sample image is identified through the training model to complete the segmentation task of stem cells in the sample image, and the cell confluence is calculated based on the segmentation results. However, this technology relies on the segmentation neural network model and requires a large amount of data. In addition, the training process of the model also requires a lot of computing resources and has high requirements for computer hardware. Even after the model training is completed, a certain amount of computing resources still needs to be consumed in the process of applying the model. Since neural networks are black box in nature, when the amount of sample data is insufficient or the data is relatively single, this technology cannot guarantee generalization.

[0031] Therefore, in view of the above-mentioned various problems, the present application embodiment provides a method for calculating cell confluence, which can use morphological image processing technology to enhance the pixel change gradient in the adherent cell image, realize the enhancement of the original image, and highlight the cell area in the image. After completing the image enhancement, the present application embodiment can realize the recognition of the cell area in the image based on the gradient change, obtain the complete cell area, and calculate the cell confluence. Compared with the aforementioned K-Means clustering algorithm, which calculates the cell confluence by iteratively calculating the distance between the sample point and the cluster center, the present application embodiment does not need to adjust the cluster center, does not need to select the K value, and reduces the factors affecting the stability of the algorithm. Compared with constructing an optical flow model and fusion features to analyze the cell confluence of mesenchymal stem cells, the algorithm complexity of the present application embodiment is lower, the algorithm efficiency is higher, and the hardware requirements are also lower. In addition, the application process of the present application embodiment does not need to construct a data set training segmentation neural network model, which is more conducive to realizing automatic, fast and accurate cell confluence calculation in various application scenarios.

[0032] The technical solution of the present application is described below through specific embodiments.

[0033] Reference Figure 1 , shows a schematic diagram of a method for calculating cell confluence provided in an embodiment of the present application, which may specifically include the following steps:

[0034] S101. Collect adherent cell images of stem cells.

[0035] It should be noted that the present method can be applied to a computer device, which can automatically, quickly and accurately calculate the cell confluence by executing the various steps of the method provided in the embodiment of the present application. That is, the execution subject of the embodiment of the present application can be a computer device, which can be any type of device with relevant functions, such as image acquisition function, image processing function and other relevant functions required to implement the present method. The embodiment of the present application does not limit the type of computer device.

[0036] In an embodiment of the present application, the adherent cell image of stem cells can be obtained by taking an image of the stem cells after the cells grow adherently and fully expand in a culture dish or culture bottle. The above-mentioned adherent cell image is the original image used to calculate the cell confluence, that is, the original adherent cell image.

[0037] In a possible implementation of the embodiment of the present application, images of adherent stem cells to be identified may be captured by a microscope at one or more time points during the cell culture process to obtain images of adherent cells.

[0038] S102, performing optimization processing on the adherent cell image to obtain a target image, wherein the optimization processing includes enhancing the cell image features in the adherent cell image.

[0039] In the embodiment of the present application, for the collected original image of adherent cells, the computer device may first perform image optimization processing on it, and perform subsequent related processing and calculation steps based on the target image obtained after optimization. The optimization processing in the embodiment of the present application may include one or more types of processing to enhance the cell image features in the original image of adherent cells.

[0040] In a possible implementation of the embodiment of the present application, the original image of adherent cells can be optimized by processing the original image based on the opening operation and closing operation in the morphological image processing algorithm. On this basis, the original image can also be optimized by combining brightness equalization means such as contrast enhancement and gamma correction to strengthen the pixel gradient change of the stem cell area image and highlight the characteristics of the stem cell image.

[0041] Exemplarily, when optimizing the original image of adherent cells, an opening operation and / or a closing operation may be used to optimize the original image of adherent cells to obtain a target image.

[0042] In image processing algorithms, the opening operation is a serial combination operation that first erodes the image and then dilates it; the closing operation is to first dilate the image and then erode it. The order of erosion and dilation in the opening and closing operations is opposite.

[0043] Among them, the processing process of the opening operation is to subtract the result of the image opening operation from the original image, so as to capture the peak information in the image. The opening operation is mainly aimed at bright objects on a dark background. It can correct the influence of uneven lighting, delete most of the non-uniform background, and separate the foreground and background of the image. The processing process of the closing operation is to subtract the result of the image closing operation from the original image, so as to capture the valley information in the image. The closing operation can highlight the dark details in the image and separate the background information that may exist in the image.

[0044] In the embodiment of the present application, in order to optimize the original image of adherent cells, the original image may be processed by an opening operation or a closing operation, or may be processed by an opening operation and a closing operation at the same time.

[0045] Exemplarily, the adherent cell image can be subjected to an opening operation and a closing operation in sequence to obtain a first image and a second image, respectively. That is, the first image is an image obtained after the computer device performs an opening operation on the adherent cell image, and the second image is an image obtained after the computer device performs a closing operation on the adherent cell image. Then, based on the original image of the adherent cell, an image obtained by subtracting the first image from the original image can be added, and an image obtained by subtracting the original image from the second image can be subtracted, thereby obtaining an optimized target image, thereby enhancing the cell region information in the original image.

[0046] The above process can be expressed as:

[0047] Target(I)=I+(I-opening(I))-(closing(I)-I)

[0048] Among them, Target(I) is the target image to be optimized, I represents the original image of adherent cells, opening(I) represents the image obtained by opening the original image of adherent cells I, i.e., the first image, and closing(I) represents the image obtained by closing the original image of adherent cells I, i.e., the second image. Therefore, I-opening(I) represents the image obtained by subtracting the first image from the original image, and closing(I)-I represents the image obtained by subtracting the original image from the second image.

[0049] In a possible implementation of the embodiment of the present application, after the original image is processed based on the opening operation and the closing operation in the morphological image processing algorithm to obtain the optimized target image, the computer device can perform subsequent processing on the target image to calculate the cell confluence of the stem cells. That is, the computer device can calculate the cell confluence of the stem cells by executing the steps shown in S103 and S104.

[0050] In another possible implementation of the embodiment of the present application, after optimizing the original image of adherent cells by using opening and / or closing operations, the computer device can further optimize the image by processing methods such as contrast enhancement and brightness balance, and use the image obtained after further optimization as the target image for subsequent processing.

[0051] In the embodiment of the present application, when contrast enhancement processing is performed on an image, contrast enhancement algorithms such as histogram equalization and linear stretching can be used but not limited to be implemented. When brightness equalization processing is performed on an image, image brightness equalization adjustment can be implemented based on gamma correction. Among them, the gamma value (γ) is a nonlinear parameter that can be used to describe the relationship between input and output. By using a power function formula to describe the relationship between the input pixel value (Iin) and the output pixel value (Iout), brightness equalization processing of the image can be implemented. The above processing process can be expressed as:

[0052] Iout=Iin γ

[0053] like Figure 2 FIG. 1 is a schematic diagram of optimizing the image of adherent cells provided in an embodiment of the present application. Figure 2 (a) shows a schematic diagram of the original image of adherent cells. Figure 2 (b) shows the above-mentioned optimization processing methods for Figure 2 (a) in FIG. 1 is a schematic diagram of the target image obtained after the original image is optimized, that is, Figure 2 (b) shows the Figure 2 The effect diagram of (a) after optimization. Figure 2 As can be seen from (a) and (b) in the figure, the image features of stem cells are clearer and more prominent in the target image after optimization. Figure 2 By performing subsequent processing and calculation on the target image shown in (b) in the figure, more accurate cell confluence results can be obtained.

[0054] S103: Determine the cell region in the target image through edge detection.

[0055] In an embodiment of the present application, after completing the optimization processing of the original image of adherent cells and obtaining a target image containing clearer and more prominent cell image features, the computer device can lock the cell area in the target image through algorithms such as edge detection to obtain a complete cell area that can be used to calculate cell confluence.

[0056] In the embodiment of the present application, the process of identifying a complete cell region from a target image may include edge detection, image connected domain calculation, and flood filling. Among them, edge detection of the target image can preliminarily identify the edge of the cell region in the image. On this basis, through image connected domain calculation and flood filling, the foreground and background of the image can be accurately identified, and the complete cell region can be screened out.

[0057] In a possible implementation of the embodiment of the present application, as Figure 3 As shown, determining the cell region in the target image by edge detection in S103 may specifically include the following steps S1031-S1033:

[0058] S1031. Determine the gradient strength of each pixel in the target image through edge detection.

[0059] In the embodiment of the present application, the Sobel operator or other similar edge detection operators may be used to calculate the gradient strength and direction of each pixel in the target image. The embodiment of the present application does not limit the specific method of calculating the gradient strength of the pixel.

[0060] S1032. Determine a plurality of continuous edge points in the target image according to the gradient strength of each pixel point, wherein the plurality of continuous edge points constitute a cell edge region in the target image.

[0061] In an embodiment of the present application, the calculated gradient strength of each pixel point can be used to determine edge points in the target image, and these edge points can constitute the cell edge region in the image.

[0062] In a possible implementation of the embodiment of the present application, multiple potential edge points in the target image can be determined based on the gradient strength of each pixel point. The gradient strength of any potential edge point is the maximum value in at least one region in the target image, that is, the potential edge point can be a pixel point with the maximum gradient strength in a certain region in the target image.

[0063] Gradient can be understood as the intensity of pixel changes. Since there are multiple edges in the actual cell image, after calculating the gradient intensity of each pixel, the computer device can take the pixel with stronger gradient intensity or the pixel with the maximum gradient intensity in a certain area as a potential edge point.

[0064] However, in the area where the edge strength is detected, not all pixels are the real edge points of the cell area. Therefore, a non-maximum suppression step is required, that is, each pixel is checked to ensure that the gradient strength of the detected edge point is the largest in its neighborhood. If the gradient strength of an edge point is not the largest in its neighborhood, it needs to be suppressed to ensure that the detected edge points are continuous.

[0065] Specifically, after detecting and obtaining multiple potential edge points, the gradient strength of each potential edge point can be compared with the gradient strength of each potential edge point in its neighborhood. By extracting each potential edge point whose gradient strength is the maximum in its neighborhood, multiple continuous edge points in the target image can be obtained.

[0066] In a possible implementation of an embodiment of the present application, in order to further remove possible false edge points, the gradient intensity can also be binarized by setting a high threshold and a low threshold. In this process, edge points above the high threshold can be considered as real edge points, while those below the low threshold are considered as noise. If the gradient intensity of a certain pixel is between the high threshold and the low threshold, it will only be retained as an edge point when it is connected to a pixel above the high threshold, otherwise the pixel will also be suppressed or removed. The setting of the above-mentioned high threshold and the low threshold can be determined according to actual needs, and the embodiment of the present application is not limited to this.

[0067] S1033. Obtain the cell region in the target image by segmenting the cell edge region.

[0068] After the above steps are used to identify the real continuous edge points in the target image, the computer device can perform segmentation and other processing on the cell edge area formed by these edge points, so as to obtain the complete cell area in the target image.

[0069] In a possible implementation of the embodiment of the present application, for the target image obtained through the processing of the aforementioned step S1032, the computer device may first fill each hole region in the image whose area is smaller than the first threshold value through a closing operation.

[0070] Specifically, for the target image obtained by processing in step S1032, the computer device can connect the edge points in the image through the expansion algorithm to obtain the cell edge area, and then fill the small area, such as the hole area with an area less than the first threshold, through the closing operation. The above expansion and closing operation processing process can be iterated multiple times according to the actual needs of the target image, and the embodiment of the present application is not limited to this.

[0071] After filling the smaller hole areas, the computer device can determine the connected domains in the target image after filling each smaller hole area, and fill each connected domain with an area smaller than the second threshold as the image background to obtain a third image. The third image only represents the image obtained after filling each connected domain with an area smaller than the second threshold as the image background, and has no other reference meaning.

[0072] Specifically, for an image after filling a smaller hole area, the computer device can calculate the connected domain in the image, and by setting a threshold, for example, by setting a second threshold, identify small connected domains in the image whose area is smaller than the second threshold, and then fill these small connected domains as the background to obtain a third image. The size of the above-mentioned second threshold can be set according to actual needs, and its specific value can be greater than the above-mentioned first threshold, or less than the first threshold, or equal to the first threshold, and the embodiment of the present application does not limit this.

[0073] In a possible implementation manner of the embodiment of the present application, the pixels of each pixel point in the identified small connected domain whose area is smaller than the second threshold value may be filled with 0 as the background.

[0074] In the embodiment of the present application, for the third image, the computer device can continue to fill the smaller hole area on this basis, and obtain the foreground in the image through relevant processing means.

[0075] Specifically, the computer device can use the flooding algorithm to fill the smaller hole areas in the third image. For example, each hole area in the third image whose area is smaller than the second threshold value can be filled, and then the connected domain in the third image is calculated after the third image is inverted, and each connected domain whose area is smaller than the third threshold value is filled as the image foreground by setting a threshold value, thereby obtaining a fourth image. Exemplarily, by setting a threshold value, the pixel value of each pixel point in each small connected domain in the image obtained after the flooding algorithm is processed can be filled with 1 as the foreground of the image, and the image obtained at this time is the fourth image.

[0076] In this way, the computer device can obtain the complete cell area in the target image based on the fourth image recognition.

[0077] In a possible implementation of the embodiment of the present application, after obtaining the fourth image, the computer device can calculate the connected domain in the fourth image after performing an inversion operation on the fourth image, and determine the boundary points of each connected domain whose area is less than the fourth threshold as the region growth seed points. Then, based on the target image, the region growth seed points are used to perform region growth to obtain the fifth image, and the complete cell region in the target image can be obtained by performing edge segmentation on the fifth image. The above-mentioned second threshold, third threshold and fourth threshold can be equal or different, and the specific numerical values ​​of these thresholds can be determined according to actual needs, and the embodiment of the present application does not limit this.

[0078] In a possible implementation of the embodiment of the present application, region growing based on region growing seed points may be performed according to the following algorithm:

[0079] (1) Scaling the original image, the scaling factor can be determined according to the actual situation of the image. The original image can be the target image obtained after image optimization processing.

[0080] (2) Formulate region growing rules based on prior knowledge. Since the background pixel variation in the cell-free region of the image is small, the boundary points of the connected domain of the background region can be locked as the seed points for region growing.

[0081] (3) The region growing step size can be set to 2, and region growing can be performed based on the eight-neighborhood criterion.

[0082] (4) Performing region growing based on pixel distance criteria: First, determine whether the neighboring points of the region growing seed point are background points with a pixel value of 0 on the fourth image obtained in the above step; then perform region growing based on two major principles. That is, if it is a background point, stop growing; if the pixel value is greater than a preset pixel threshold, stop growing, and assign the pixel that meets the conditions a value of 0 as a background point.

[0083] like Figure 4 FIG. 1 is a schematic diagram of a target image segmentation process provided by an embodiment of the present application. Figure 4 (a) shows the target image after optimization. Figure 4 The target image shown in (a) is segmented to distinguish the foreground and background of the image. Figure 4 As shown in (b), Figure 4 (b) shows the fourth image in the above step. Figure 4 After region growing, the image shown in (b) is obtained as follows Figure 4 As shown in (c) in .

[0084] for Figure 4 In the image shown in (c), the computer device can use bilateral filtering, smoothing and segmentation and other processing methods to obtain the edges of each cell region in the image, thereby determining the complete cell region in the image for calculation of cell confluence in the next step.

[0085] In a possible implementation of the embodiment of the present application, after completing the region growth to obtain complete cell regions, the computer device can also obtain the contour points of the cell region in the image, perform curve fitting and other processing on the contour points, and beautify the contour of the cell region. The above-mentioned curve fitting can be implemented using algorithms including but not limited to multi-term fitting, Bezier curve fitting, etc., and the embodiment of the present application does not limit this.

[0086] S104: Calculate the cell confluence of the stem cells based on the cell region in the target image.

[0087] In an embodiment of the present application, the computer device can calculate the accurate cell confluence on a pixel level based on the complete cell area obtained by segmentation.

[0088] Specifically, the computer device can calculate the total number of pixels in the target image and the total number of pixels in each cell region of the target image, thereby calculating the cell confluence of the stem cells based on the total number of pixels in the target image and the total number of pixels in each cell region. The above calculation process can be expressed as:

[0089]

[0090] That is, the cell confluence of the stem cells can be obtained by dividing the calculated total number of pixels in each adherent cell region in the target image by the total number of all pixels in the target image.

[0091] The embodiment of the present application collects the adherent cell image of the stem cell, and can determine the cell area in the target image by edge detection on the basis of optimizing the adherent cell image to obtain the target image, so as to calculate the cell confluence of the stem cell based on the cell area in the target image. Compared with the artificial observation method used in the prior art to evaluate the cell confluence, the calculation method evaluation standard used in the embodiment of the present application is more objective and more capable of large-scale production operations. Secondly, the relevant algorithms and processing steps used in the embodiment of the present application are more capable of accurately detecting the complete cell area compared to the simple and fast threshold segmentation in the prior art; compared to the K-Means clustering algorithm with more complex logic, the embodiment of the present application needs to adjust fewer parameters during the application process, the algorithm effect is more stable, and the algorithm efficiency is also higher. Third, the application of the embodiment of the present application does not need to pre-make a data set, which can save a lot of manpower costs, and the calculation process also requires too much computing resources, and the requirements for equipment hardware conditions are relatively low.

[0092] It should be noted that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0093] For ease of understanding, the following is an introduction to the method for calculating cell confluence provided in the embodiment of the present application in combination with a complete example. The method mainly includes steps such as image acquisition, image optimization, image segmentation and confluence calculation. Specifically:

[0094] 1. In the image acquisition step, the adherent stem cells to be identified can be acquired as original images of the adherent cells at one or more time points during the cell culture process using a microscope.

[0095] 2. In the image optimization step, the original image is processed mainly based on the opening and closing operations in the morphological image processing algorithm, and then the original image is optimized by combining contrast enhancement, gamma correction and other brightness balancing methods to strengthen the pixel gradient changes of the stem cell area image and highlight the stem cell image characteristics. The relevant steps include:

[0096] Step 2.1, the opening operation is a serial combination operation of erosion followed by dilation. The result of the image opening operation can be subtracted from the original image to capture the peak information in the image. The opening operation is mainly used for bright objects on a dark background to correct the influence of uneven lighting. It can delete most of the non-uniform background and achieve separation of foreground and background.

[0097] Step 2.2: The closing operation is to first dilate the image and then perform the erosion operation. The result after the image closing operation is subtracted from the original image can capture the valley information in the image, highlight the dark details in the image, and separate the possible background information in the image.

[0098] Step 2.3: Add the image in step 2.1 to the original image and subtract the image in step 2.2 to enhance the cell region information. The above process can be expressed as:

[0099] Target(I)=I+(I-opening(I))-(closing(I)-I)

[0100] Wherein, Target(I) is the target image to be optimized, I represents the original image of adherent cells, opening(I) represents the image obtained by opening the original image of adherent cells I, i.e., the first image, and closing(I) represents the image obtained by closing the original image of adherent cells I, i.e., the second image.

[0101] Step 2.4: Contrast enhancement: The image may be processed using, but not limited to, contrast enhancement algorithms such as histogram equalization and linear stretching.

[0102] Step 2.5, brightness balance. Image brightness balance can be adjusted based on gamma correction. The gamma value is a nonlinear parameter used to describe the relationship between input and output. A power function formula can be used to describe the relationship between the input pixel value (Iin) and the output pixel value (Iout). γ is the gamma value. The relevant formula can be expressed as follows:

[0103] Iout=Iin γ

[0104] Figure 2 The schematic diagram of optimizing the image of adherent cells using the above steps 2.1-2.5 is shown. Figure 2 (a) shows the original image of adherent cells. Figure 2 (b) shows the above steps 2.1-2.5 for Figure 2 (a) is a diagram showing the effect of the target image obtained after the original image is optimized.

[0105] 3. In the image segmentation step, the edge detection algorithm can be used to lock the cell area, and the complete cell area can be obtained through algorithms such as image connected domain calculation and flood filling. The relevant steps include:

[0106] Step 3.1: Sobel operator or other similar edge detection operators can be used to calculate the gradient strength and direction of each pixel in the target image. By determining the location of the pixels with the strongest gradient strength change or the maximum gradient strength change in a certain area in the target image, these pixels can be identified as potential edge points.

[0107] Step 3.2: In the area where the edge strength is detected, not all pixels are true edge points. Therefore, a non-maximum suppression step is required, that is, each edge point is checked to ensure that the gradient strength of the point is the maximum value in its neighborhood. If not, the pixel can be suppressed to ensure that the detected edge points are continuous.

[0108] Step 3.3: To further remove false edge points, the gradient intensity can be binarized by setting a high threshold and a low threshold. Edge points with gradient intensity higher than the high threshold can be considered as true edge points, while those below the low threshold are considered as noise. If the gradient intensity of a pixel is between the high threshold and the low threshold, it will be retained as an edge point only if it is connected to a pixel with a gradient intensity higher than the high threshold.

[0109] Step 3.4, connect the cell edge area obtained by executing step 3.3 through the expansion algorithm, and fill the smaller holes therein through the closing operation. The above-mentioned smaller holes can be the hole areas whose area is smaller than the aforementioned first threshold. In this process, multiple iterations of the expansion algorithm and the closing operation can be performed according to the actual needs of the image.

[0110] Step 3.5: Based on the binary image obtained in step 3.4, by calculating the connected domain, a threshold value, such as the aforementioned second threshold value, can be set, and the pixels in each small connected domain independent of the large connected domain are filled with 0 as the background. Among them, the large connected domain may refer to a connected domain with an area greater than the aforementioned second threshold value, and the small connected domain may refer to a connected domain with an area less than the second threshold value. The image obtained in this step is the third image in the aforementioned embodiment.

[0111] Step 3.6: Based on the binary image obtained in step 3.5, i.e., the third image, the flooding algorithm can be used to fill the smaller holes first, and after the image inversion operation, the connected domain in the image is calculated, and by setting a threshold value, such as the third threshold value, the pixel value of each pixel point in the small connected domain of the image obtained after the flooding algorithm is filled with 1 as the foreground of the image. The image obtained in this step is the fourth image in the aforementioned embodiment.

[0112] Step 3.7, based on the binary image obtained in step 3.6, i.e., the fourth image, the image is inverted again to calculate the connected domain in the image, and by setting a threshold, such as the fourth threshold, the connected domain boundary point combination with an area less than the fourth threshold is regarded as the region growing seed point. In this way, based on the original image, the region growing seed point can be used to perform region growing, and the original image for region growing can be the target image after optimization processing. The region growing algorithm can be as follows:

[0113] (1) Scaling the original image, the scaling factor can be determined according to the actual situation of the image. The original image can be the target image obtained after image optimization processing.

[0114] (2) Formulate region growing rules based on prior knowledge. Since the background pixel variation in the cell-free region of the image is small, the boundary points of the connected domain of the background region can be locked as the seed points for region growing.

[0115] (3) The region growing step size can be set to 2, and region growing can be performed based on the eight-neighborhood criterion.

[0116] (4) Performing region growing based on pixel distance criteria: First, determine whether the neighboring points of the region growing seed point are background points with a pixel value of 0 on the fourth image obtained in the above step; then perform region growing based on two major principles. That is, if it is a background point, stop growing; if the pixel value is greater than a preset pixel threshold, stop growing, and assign the pixel that meets the conditions a value of 0 as a background point.

[0117] Figure 4 The schematic diagram of processing the target image using the above steps 3.1-3.7 is shown. Figure 4 (a) shows the target image after optimization. Figure 4 (b) shows the binary image obtained after the processing in the above step 3.6, i.e., the fourth image. Figure 4 (c) in the figure shows the image obtained after region growing.

[0118] Step 3.8: Use bilateral filtering to smooth the edges of the segmentation results to obtain the complete cell area in the image.

[0119] 4. In the confluence calculation step, the cell confluence can be calculated on a pixel level basis based on the clinical cell confluence calculation formula. Specifically, based on the processing of the above steps, the adherent cell area information can be obtained. By calculating the total number of pixels in the adherent cell area and the total number of all pixels in the image, the accurate cell confluence can be obtained by dividing the former by the latter. The above calculation process can be expressed as:

[0120]

[0121] The embodiment of the present application processes the original image of adherent cells by utilizing the open operation, closed operation, etc. in the morphological image processing algorithm, and then optimizes the original image by combining contrast enhancement, gamma correction and other brightness equalization means, which can strengthen the pixel gradient change of the stem cell area image and highlight the characteristics of the stem cell image. Then, the cell area is locked by utilizing the edge detection algorithm, and the complete cell area can be obtained by algorithms such as image connected domain calculation, flood filling and optimized regional growth. On this basis, stem cell image cell area segmentation and cell confluence calculation can be quickly realized based on traditional algorithms. The calculation method of cell confluence provided by the application embodiment can quickly and accurately realize the calculation of cell confluence.

[0122] Reference Figure 5 , shows a schematic diagram of a cell confluence calculation device provided in an embodiment of the present application, which may specifically include an image acquisition module 501, an image optimization module 502, an image segmentation module 503 and a cell confluence calculation module 504, wherein:

[0123] An image acquisition module 501 is used to acquire images of adherent cells of stem cells;

[0124] An image optimization module 502 is used to optimize the adherent cell image to obtain a target image, wherein the optimization includes enhancing the cell image features in the adherent cell image;

[0125] An image segmentation module 503 is used to determine the cell area in the target image by edge detection;

[0126] The cell confluence calculation module 504 is used to calculate the cell confluence of the stem cells based on the cell region in the target image.

[0127] In a possible implementation of the embodiment of the present application, the image optimization module 502 may be specifically configured to: optimize the adherent cell image by using an opening operation and / or a closing operation to obtain a target image.

[0128] Exemplarily, the image optimization module 502 can be used to: perform opening and closing operations on the adherent cell image in sequence to obtain a first image and a second image, respectively; based on the adherent cell image, add an image obtained by subtracting the first image from the adherent cell image, and subtract an image obtained by subtracting the adherent cell image from the second image to obtain the target image.

[0129] In another possible implementation of the embodiment of the present application, the image segmentation module 503 can be specifically used to: determine the gradient strength of each pixel in the target image through edge detection; determine a plurality of continuous edge points in the target image based on the gradient strength of each pixel, and the plurality of continuous edge points constitute a cell edge region in the target image; and obtain a cell region in the target image by segmenting the cell edge region.

[0130] In one example, the image segmentation module 503 can be used to: determine multiple potential edge points in the target image based on the gradient strength of each pixel point, the gradient strength of any potential edge point being the maximum value in at least one area in the target image; compare the gradient strength of each potential edge point with the gradient strength of each potential edge point in its neighborhood; extract each potential edge point whose gradient strength is the maximum value in its neighborhood, and obtain multiple continuous edge points in the target image.

[0131] In another example, the image segmentation module 503 can also be used to: fill each hole area whose area is smaller than a first threshold value through a closing operation; determine the connected domain in the target image after filling each of the hole areas, and fill each connected domain whose area is smaller than a second threshold value as the image background to obtain a third image; calculate the connected domain in the third image after performing a negation operation on the third image, and fill each connected domain whose area is smaller than the third threshold value as the image foreground to obtain a fourth image; and identify the cell area in the target image based on the fourth image.

[0132] In another example, the image segmentation module 503 can also be used to: calculate the connected domains in the fourth image after performing a negation operation on the fourth image, and determine the boundary points of each connected domain whose area is less than a fourth threshold as region growth seed points; based on the target image, perform region growth using the region growth seed points to obtain a fifth image; perform edge segmentation on the fifth image to obtain a cell region in the target image.

[0133] In another possible implementation of the embodiment of the present application, the cell confluence calculation module 504 can be specifically used to: respectively calculate the total number of pixels in the target image and the total number of pixels in the cell area; and calculate the cell confluence of the stem cells based on the total number of pixels in the target image and the total number of pixels in the cell area.

[0134] An embodiment of the present application provides a device for calculating cell confluence. By using the device, each step in the aforementioned method embodiments can be implemented.

[0135] As for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method embodiment part.

[0136] Reference Figure 6 , shows a schematic diagram of a computer device provided by an embodiment of the present application. Figure 6 As shown, the computer device 600 in the embodiment of the present application includes: a processor 610, a memory 620, and a computer program 621 stored in the memory 620 and executable on the processor 610. When the processor 610 executes the computer program 621, the steps in each embodiment of the above-mentioned method for calculating cell confluence are implemented, for example Figure 1 Alternatively, when the processor 610 executes the computer program 621, the functions of each module / unit in the above-mentioned device embodiments are realized, for example Figure 4 Functions of modules 401 to 404 are shown.

[0137] Exemplarily, the computer program 621 may be divided into one or more modules / units, which are stored in the memory 620 and executed by the processor 610 to complete the present application. The one or more modules / units may be a series of computer program instruction segments capable of completing specific functions, which may be used to describe the execution process of the computer program 621 in the computer device 600. For example, the computer program 621 may be divided into an image acquisition module, an image optimization module, an image segmentation module, and a cell confluence calculation module, and the specific functions of each module are as follows:

[0138] An image acquisition module, used for acquiring adherent cell images of stem cells;

[0139] An image optimization module, used for optimizing the adherent cell image to obtain a target image, wherein the optimization includes enhancing the cell image features in the adherent cell image;

[0140] An image segmentation module, used for determining the cell area in the target image by edge detection;

[0141] The cell confluence calculation module is used to calculate the cell confluence of the stem cells based on the cell region in the target image.

[0142] The computer device 600 may be a device capable of implementing each step in the above-mentioned method embodiments, and may be a computing device such as a desktop computer or a cloud server. The computer device 600 may include, but is not limited to, a processor 610 and a memory 620. Those skilled in the art may understand that Figure 6This is only an example of the computer device 600 and does not constitute a limitation of the computer device 600. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the computer device 600 may also include input and output devices, network access devices, buses, etc.

[0143] The processor 610 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0144] The memory 620 may be an internal storage unit of the computer device 600, such as a hard disk or memory of the computer device 600. The memory 620 may also be an external storage device of the computer device 600, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the computer device 600. Further, the memory 620 may also include both an internal storage unit of the computer device 600 and an external storage device. The memory 620 is used to store the computer program 621 and other programs and data required by the computer device 600. The memory 620 may also be used to temporarily store data that has been output or is to be output.

[0145] An embodiment of the present application also discloses a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method for calculating cell confluence as described in the aforementioned embodiments is implemented.

[0146] The embodiments of the present application further disclose a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for calculating the cell confluence as described in the above embodiments is implemented.

[0147] The embodiments of the present application further disclose a computer program product. When the computer program product is run on a computer, the computer is enabled to execute the method for calculating cell confluence described in the aforementioned embodiments.

[0148] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application is described in detail with reference to the above-mentioned embodiments, a person skilled in the art should understand that the technical solutions described in the above-mentioned embodiments can still be modified, or some of the technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A method for calculating cell confluence, characterized in that: include: Collect adherent cell images of stem cells; Performing optimization processing on the adherent cell image to obtain a target image, wherein the optimization processing includes enhancing the cell image features in the adherent cell image; Determine the cell area in the target image by edge detection; The cell confluence of the stem cells is calculated based on the cell area in the target image.

2. The method according to claim 1, characterized in that: The step of optimizing the adherent cell image to obtain a target image comprises: The adherent cell image is optimized by using an opening operation and / or a closing operation to obtain a target image.

3. The method according to claim 2, characterized in that The method of optimizing the adherent cell image by using an opening operation and / or a closing operation to obtain a target image includes: performing an opening operation and a closing operation on the adherent cell image in sequence to obtain a first image and a second image respectively; Based on the adherent cell image, an image obtained by subtracting the first image from the adherent cell image is added, and an image obtained by subtracting the adherent cell image from the second image is subtracted to obtain the target image.

4. The method according to any one of claims 1 to 3, characterized in that: The step of determining the cell region in the target image by edge detection comprises: Determine the gradient strength of each pixel in the target image by edge detection; Determine a plurality of continuous edge points in the target image according to the gradient strength of each pixel point, wherein the plurality of continuous edge points constitute a cell edge region in the target image; The cell region in the target image is obtained by segmenting the cell edge region.

5. The method according to claim 4, characterized in that Determining a plurality of continuous edge points in the target image according to the gradient strength of each pixel point includes: Determine a plurality of potential edge points in the target image according to the gradient strength of each pixel point, wherein the gradient strength of any potential edge point is a maximum value in at least one region in the target image; Compare the gradient strength of each potential edge point with the gradient strength of each potential edge point in its neighborhood; Each potential edge point whose gradient intensity is the maximum value in its neighborhood is extracted to obtain a plurality of continuous edge points in the target image.

6. The method according to claim 4, characterized in that The step of obtaining the cell region in the target image by segmenting the cell edge region comprises: Fill each hole region whose area is smaller than the first threshold value by a closing operation; Determine a connected domain in the target image after filling each of the hole regions, and fill each connected domain whose area is smaller than the second threshold as the image background to obtain a third image; After performing a negation operation on the third image, a connected domain in the third image is calculated, and each connected domain whose area is smaller than a third threshold is filled as an image foreground, so as to obtain a fourth image; A cell region in the target image is identified based on the fourth image.

7. The method according to claim 6, characterized in that The identifying a cell region in the target image based on the fourth image comprises: After performing a negation operation on the fourth image, a connected domain in the fourth image is calculated, and boundary points of each connected domain whose area is smaller than a fourth threshold are determined as region growth seed points; Based on the target image, performing region growing using the region growing seed points to obtain a fifth image; Perform edge segmentation on the fifth image to obtain a cell region in the target image.

8. The method according to any one of claims 1-3 or 5-7, characterized in that: The step of calculating the cell confluence of the stem cells based on the cell region in the target image comprises: Calculating the total number of pixels in the target image and the total number of pixels in the cell area respectively; The cell confluence of the stem cells is calculated according to the total number of pixels in the target image and the total number of pixels in the cell region.

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, the method for calculating cell confluence according to any one of claims 1 to 8 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for calculating cell confluence according to any one of claims 1 to 8 is implemented.

Images