Stem cell differentiation degree recognition method based on image features

Through image feature analysis, the structural and morphological differentiation factors of stem cells were calculated, which solved the problem of identification difficulties caused by the complex image features during stem cell differentiation, and achieved an accurate assessment of the degree of stem cell differentiation.

CN118710646BActive Publication Date: 2024-11-15SHAANXI HEZE NORTHWEST BIOTECHNOLOGY CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411194582.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-29
Publication Date
2024-11-15
Estimated Expiration
2044-08-29

AI Technical Summary

Technical Problem

The prior art cannot accurately identify the degree of stem cell differentiation, mainly due to the complex internal structure and overall morphological changes of stem cells, resulting in complex image features and difficult to accurately identify.

Method used

The stem cell differentiation degree recognition method based on image characteristics is used to obtain the grayscale image of stem cells, perform edge detection and threshold segmentation, calculate structural differentiation factors and morphological differentiation factors, and combine these factors to calculate the differentiation evaluation value to evaluate the degree of stem cells.

Benefits of technology

Through structural and morphological differentiation factors, the degree of differentiation of stem cells is accurately identified and evaluated, and the difficulty of identification caused by the complex image characteristics during stem cell differentiation in the prior art is solved, and the accuracy of identification and evaluation of the degree of differentiation of stem cells is achieved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118710646B_ABST
    Figure CN118710646B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of image analysis, and specifically to a method for identifying the degree of differentiation of stem cells based on image features. The method first performs edge detection on the grayscale image of the stem cell to be tested, and obtains the stem cell region and the edge lines in the stem cell region; obtains the cytoplasm region and the nucleus region by threshold segmentation; obtains the structural differentiation factor according to the grayscale value and number of pixels in the cytoplasm region and the nucleus region, the position distribution of the edge pixels in the nucleus region, and the number of edge lines in the stem cell region; performs ellipse fitting on the edge of the stem cell region to obtain the ellipse-like coefficient of the stem cell region; obtains the morphological differentiation factor according to the grayscale value change between the edge pixels and the centroid of the stem cell region and the ellipse-like coefficient; and evaluates the degree of differentiation of the stem cell to be tested based on the structural differentiation factor and the morphological differentiation factor. The present invention can accurately identify and evaluate the degree of differentiation of stem cells.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of image analysis, and in particular to a method for identifying the degree of stem cell differentiation based on image features. Background Art

[0002] Stem cells have the ability to self-renew and differentiate into a variety of cell types. During the differentiation process of stem cells, they will gradually develop towards mature functional cells such as red blood cells. Identifying and evaluating the degree of stem cell differentiation into red blood cells is crucial for monitoring the quality of stem cell therapy, predicting its efficacy, and understanding related disease mechanisms.

[0003] In related technologies, the degree of differentiation of stem cells is usually identified and evaluated based on the image features of the internal structure and overall morphology of stem cells. However, during the differentiation process of stem cells, the changes in their internal structure and overall morphology are relatively complex, which makes the image features of stem cells also relatively complex, which in turn makes it impossible for existing methods to accurately identify the degree of differentiation of stem cells. Summary of the invention

[0004] In order to solve the technical problem that the changes in the internal structure and overall morphology of stem cells during differentiation are relatively complex, resulting in complex image features of stem cells, and thus the existing methods cannot accurately identify the degree of differentiation of stem cells, the purpose of the present invention is to provide a method for identifying the degree of differentiation of stem cells based on image features. The technical solution adopted is as follows:

[0005] The present invention proposes a method for identifying the degree of stem cell differentiation based on image features, the method comprising:

[0006] Acquire a grayscale image of the stem cells to be tested; perform edge detection on the grayscale image to obtain a stem cell region and an edge line in the stem cell region;

[0007] Performing threshold segmentation on the stem cell region to obtain a cytoplasm region and a cell nucleus region; obtaining a structural differentiation factor of the stem cell to be tested according to the grayscale value and number of pixel points in the cytoplasm region and the cell nucleus region, the position distribution of edge pixel points in the cell nucleus region, and the number of edge lines in the stem cell region;

[0008] Performing ellipse fitting on the edge of the stem cell region to obtain a quasi-ellipse coefficient of the stem cell region; obtaining a morphological differentiation factor of the stem cell to be tested based on the change in the gray value of the pixel point between each edge pixel point and the centroid of the stem cell region and the quasi-ellipse coefficient; obtaining a differentiation evaluation value of the stem cell to be tested based on the structural differentiation factor and the morphological differentiation factor;

[0009] Based on the differentiation evaluation value, the differentiation degree of the stem cells to be tested is evaluated.

[0010] Furthermore, the step of obtaining the structural differentiation factor of the stem cell to be tested comprises:

[0011] Obtaining an eight-chain code sequence of edge pixels of a cell nucleus region, and obtaining a non-circular degree of the cell nucleus region according to differences between adjacent elements in the eight-chain code sequence;

[0012] Obtaining a grayscale distribution characteristic value and a quantity distribution characteristic value of the stem cell region, wherein the grayscale distribution characteristic value is positively correlated with an average grayscale value of all pixels in the cytoplasm region, the grayscale distribution characteristic value is negatively correlated with an average grayscale value of all pixels in the cell nucleus region, the quantity distribution characteristic value is positively correlated with the number of all pixels in the cytoplasm region, and the quantity distribution characteristic value is negatively correlated with the number of all pixels in the cell nucleus region;

[0013] Negative correlation mapping is performed on the number of edge lines in the stem cell area to obtain the characteristic value of the number of organelles in the stem cell area;

[0014] The non-circularity degree, the grayscale distribution characteristic value, the quantity distribution characteristic value and the organelle quantity characteristic value are integrated and normalized to obtain the structural differentiation factor of the stem cell to be tested.

[0015] Furthermore, obtaining the non-circularity degree of the cell nucleus region according to the difference between adjacent elements in the octa-chain code sequence includes:

[0016] Any two adjacent elements in the eight-chain code sequence are regarded as an element group, and the absolute value of the difference between the two elements in each element group is regarded as the element difference value of each element group;

[0017] After analyzing the discrete degrees of the element difference values ​​of all element groups and performing negative correlation mapping, the non-circular degree of the cell nucleus region is obtained.

[0018] Furthermore, performing ellipse fitting on the edge of the stem cell region to obtain the ellipse-like coefficient of the stem cell region includes:

[0019] Using the least square method, an ellipse is fitted to the edge of the stem cell region to obtain the fitting loss value of the stem cell region and the major axis and minor axis of the fitting ellipse;

[0020] The flatness coefficient of the stem cell region is obtained according to the difference between the number of pixels on the major axis and the number of pixels on the minor axis of the fitted ellipse;

[0021] Performing negative correlation mapping on the fitting loss value to obtain the ellipse similarity of the stem cell region;

[0022] The flatness coefficient and the ellipse similarity are integrated to obtain the ellipse-like coefficient of the stem cell region.

[0023] Furthermore, the step of obtaining the morphological differentiation factors of the stem cells to be tested comprises:

[0024] The line between each edge pixel point and the centroid of the stem cell region is used as a reference line segment;

[0025] Obtaining a first morphological characteristic value of the stem cell region according to the difference in grayscale value changes of the pixels on each reference line segment and the ellipse-like coefficient;

[0026] The pixel point with the maximum gray value on each reference line segment is taken as the gray turning point of each reference line segment, the line segment between the centroid and the gray turning point is taken as the inner line segment of each reference line segment, and the line segment between the edge pixel point and the gray turning point is taken as the outer line segment of each reference line segment;

[0027] Obtaining a second morphological characteristic value of the stem cell region according to the number of pixels on the inner and outer segments of each reference line segment and the change in the grayscale value of the pixels on the inner segment;

[0028] The first morphological characteristic value and the second morphological characteristic value are integrated and normalized to obtain the morphological differentiation factor of the stem cell to be tested.

[0029] Furthermore, obtaining the first morphological characteristic value of the stem cell region includes:

[0030] Taking any reference line segment as the target reference line segment, taking the absolute value of the difference between the grayscale value of each pixel point on the target reference line segment and the next adjacent pixel point as the grayscale change value of each pixel point, analyzing the overall level of the grayscale change values ​​of all pixels on the target reference line segment, and obtaining the grayscale change characteristic value of the target reference line segment;

[0031] Take any two reference line segments as reference line segment groups, and take the absolute value of the difference between the grayscale change feature values ​​of the two reference line segments in each reference line segment group as the initial grayscale change difference of each reference line segment group; perform negative correlation normalization processing on the angle between the two reference line segments in each reference line segment group to obtain the adjustment weight of each reference line segment group, and use the adjustment weight to weight the initial grayscale change difference to obtain the real grayscale change difference of each reference line segment group, where the value range of the angle is ;

[0032] The real grayscale change differences of all reference line segment groups are integrated to obtain the integrated grayscale change differences of the stem cell region;

[0033] The comprehensive grayscale change difference and the ellipse-like coefficient are integrated and normalized to obtain a first morphological characteristic value of the stem cell region.

[0034] Furthermore, obtaining the second morphological characteristic value of the stem cell region includes:

[0035] Analyze the overall level of the grayscale change values ​​of all pixels on the inner line segment of each reference line segment to obtain a first concave parameter of each reference line segment;

[0036] The number of pixels on the inner line segment of each reference line segment is used as the numerator, the number of pixels on the outer line segment of each reference line segment is used as the denominator, and the ratio is used as the second concave parameter of each reference line segment;

[0037] The first concave parameter and the second concave parameter are integrated to obtain a comprehensive concave parameter of each reference line segment; the overall level of the comprehensive concave parameters of all reference line segments is analyzed and normalized to obtain a second morphological characteristic value of the stem cell region.

[0038] Furthermore, obtaining the differentiation evaluation value of the stem cell to be tested based on the structural differentiation factor and the morphological differentiation factor comprises:

[0039] The sum of the structural differentiation factor and the morphological differentiation factor is normalized to obtain a differentiation evaluation value of the stem cell to be tested.

[0040] Furthermore, the step of evaluating the degree of differentiation of the stem cells to be tested based on the differentiation evaluation value includes:

[0041] If the differentiation evaluation value is less than the preset first threshold, the degree of differentiation of the stem cells to be tested is low; if the differentiation evaluation value is not less than the preset first threshold and is less than the preset second threshold, the degree of differentiation of the stem cells to be tested is medium; if the differentiation evaluation value is not less than the preset second threshold, the degree of differentiation of the stem cells to be tested is high.

[0042] Furthermore, the value range of the preset first threshold is , the value range of the preset second threshold is .

[0043] The present invention has the following beneficial effects:

[0044] Since the changes in the internal structure and overall morphology of stem cells during differentiation are relatively complex, the image features of stem cells are also relatively complex, which makes it impossible for existing methods to accurately identify the degree of differentiation of stem cells. The present invention first performs edge detection on the grayscale image to obtain the stem cell area and the edge line in the stem cell area. The edge line in the stem cell area is used to represent the organelles in the stem cell. In the process of stem cell differentiation into red blood cells, the volume of the cell nucleus of the stem cell will gradually decrease until it disappears, the proportion of the cytoplasm will gradually increase, and the color will become lighter, then the gray value and number of the pixel points of the cytoplasm will gradually increase, the proportion of the cell nucleus will gradually decrease, and the chromatin in the cell nucleus will become more aggregated and the color will darken. The gray value and number of pixels will gradually decrease, and the cell nucleus will gradually change from a circular shape to an irregular shape. At the same time, the stem cells to be tested will gradually lose their organelles. Therefore, the structural differentiation factor can be used to reflect the characteristics of the internal structure as the stem cell differentiation proceeds. In the process of stem cell differentiation into red blood cells, its overall morphology will gradually change to a disc shape, and a depression will appear on its inner side. In the image, the edge shape of the stem cell area is closer to an ellipse, and the gray value changes on the inner and outer sides of the stem cells are quite different. Therefore, the morphological differentiation factor can be used to reflect the characteristics of the overall morphology as the stem cell differentiation proceeds, and then the differentiation evaluation value obtained can be used to accurately identify and evaluate the differentiation degree of the stem cells to be tested. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0046] Figure 1 A flow chart of a method for identifying the degree of differentiation of stem cells based on image features provided by one embodiment of the present invention;

[0047] Figure 2 A flow chart of a method for obtaining structural differentiation factors of stem cells to be tested provided in one embodiment of the present invention;

[0048] Figure 3 This is a flow chart of a method for obtaining morphological differentiation factors of stem cells to be tested provided in one embodiment of the present invention. DETAILED DESCRIPTION

[0049] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation method, structure, features and effects of a stem cell differentiation degree identification method based on image features proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0050] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0051] The following is a detailed description of a method for identifying the degree of differentiation of stem cells based on image features provided by the present invention in conjunction with the accompanying drawings.

[0052] See also Figure 1 , which shows a flow chart of a method for identifying the degree of differentiation of stem cells based on image features provided by an embodiment of the present invention, the method comprising:

[0053] Step S1: obtaining a grayscale image of the stem cells to be tested; performing edge detection on the grayscale image to obtain a stem cell region and an edge line in the stem cell region.

[0054] Stem cells have the ability to self-renew and differentiate into a variety of cell types. During the differentiation process of stem cells, they will gradually develop in the direction of mature functional cells such as red blood cells. Identifying and evaluating the degree of differentiation of stem cells into red blood cells is crucial for monitoring the quality of stem cell therapy, predicting its efficacy, and understanding related disease mechanisms. In related technologies, the degree of differentiation of stem cells is usually identified and evaluated based on the image features of the internal structure and overall morphology of stem cells. However, during the differentiation process of stem cells, the changes in their internal structure and overall morphology are relatively complex, which leads to the image features of stem cells being relatively complex, which in turn makes the existing methods unable to accurately identify the degree of differentiation of stem cells. Therefore, an embodiment of the present invention proposes a method for identifying the degree of differentiation of stem cells based on image features to solve this problem.

[0055] In the embodiment of the present invention, a microscope imaging device is first used to collect an original image of a single stem cell to be tested in a culture dish. Since the original image is usually a multi-channel RGB image, in order to reduce the amount of calculation of subsequent image processing and improve the processing speed, in one embodiment of the present invention, the collected original image is grayed and converted into a single-channel gray image. It should be noted that graying is a technical means well known to those skilled in the art and will not be described in detail here.

[0056] Since there is still a background area in the grayscale image, it is not conducive to subsequent analysis. The grayscale value difference between the area where the stem cells to be tested are located and the background area is large, and there is an obvious edge. Therefore, the embodiment of the present invention also needs to perform edge detection on the grayscale image to obtain the stem cell area and the edge lines in the stem cell area. Among them, the organelles of the stem cells also have relatively obvious edges. Therefore, the edge lines in the stem cell area represent the edges of the organelles inside the stem cells to be tested. Subsequently, the number of organelles inside the stem cells to be tested can be used to analyze the number of edge lines in the stem cell area. Among them, edge detection is a technical means well known to those skilled in the art, and existing Canny edge detection algorithms can be used, which will not be repeated here.

[0057] It should be noted that, before culturing the test stem cells, the embodiments of the present invention need to use different colors to dye the cytoplasm and nucleus of the test stem cells, so as to facilitate the subsequent analysis of the internal image features of the test stem cells during differentiation.

[0058] Step S2: Perform threshold segmentation on the stem cell region to obtain the cytoplasm region and the nucleus region; obtain the structural differentiation factor of the stem cell to be tested based on the grayscale value and number of pixels in the cytoplasm region and the nucleus region, the position distribution of edge pixels in the nucleus region, and the number of edge lines in the stem cell region.

[0059] During the differentiation of stem cells into red blood cells, the cytoplasm and nucleus in the stem cells will undergo different changes. Since different colors are used to dye the cytoplasm and nucleus during stem cell culture, the grayscale values ​​of the cytoplasm and nucleus are quite different. Therefore, the embodiment of the present invention first performs threshold segmentation on the stem cell region to obtain the cytoplasm region and the nucleus region in the stem cell region. Among them, threshold segmentation is a technical means well known to those skilled in the art, and existing Otsu threshold segmentation algorithm can be used, which will not be described in detail here.

[0060] In the process of stem cell differentiation into red blood cells, the nucleus of the stem cell will gradually become smaller until it disappears, and the proportion of the cytoplasm will gradually increase, causing the color of the cytoplasm to become lighter, which is manifested in the grayscale image as the grayscale value and number of pixels in the cytoplasm area will gradually increase, while the proportion of the nucleus will gradually decrease, and the chromatin in the nucleus will become more aggregated, causing the color of the nucleus to gradually deepen, which is manifested in the grayscale image as the grayscale value and number of pixels in the nucleus area will gradually decrease, and the nucleus will gradually change from a circular shape to an irregular shape, and the stem cell to be tested will gradually lose its organelles. Therefore, the grayscale value and number of pixels in the cytoplasm area and the nucleus area, the position distribution of edge pixels in the nucleus area, and the number of edge lines in the stem cell area can be analyzed. The obtained structural differentiation factor reflects the characteristics of its internal structure as the differentiation of stem cells proceeds. The larger the structural differentiation factor, the greater the degree of differentiation of the stem cell to be tested.

[0061] Preferably, in one embodiment of the present invention, the method for obtaining the structural differentiation factor of the stem cell to be tested specifically includes:

[0062] See also Figure 2 , which shows a flow chart of a method for obtaining structural differentiation factors of stem cells to be tested provided in one embodiment of the present invention.

[0063] Step S201: Obtain an eight-chain code sequence of edge pixels in the cell nucleus region, and obtain the non-circularity degree of the cell nucleus region according to the difference between adjacent elements in the eight-chain code sequence.

[0064] Through the above analysis, it can be seen that as the differentiation of the cell nucleus of the stem cell proceeds, the shape of its edge will change from a circular shape to an irregular shape that is significantly different from a circle, and the chain code can reflect the shape characteristics of the edge. Therefore, the eight-chain code sequence of the edge pixel points of the cell nucleus region is first obtained, wherein the eight-chain code sequence is a technical means well known to those skilled in the art and will not be elaborated here. When the edge of the cell nucleus region is approximately circular, the difference between adjacent elements in the eight-chain code sequence is small, otherwise, the difference between adjacent elements in the eight-chain code sequence is large, so the difference between adjacent elements in the eight-chain code sequence can be analyzed to obtain the non-circular degree of the cell nucleus region. The greater the non-circular degree, the greater the difference between the edge shape of the cell nucleus region and the circle, which further indicates that the degree of differentiation of the stem cell to be tested is higher.

[0065] Preferably, in one embodiment of the present invention, the method for obtaining the non-circularity degree of the cell nucleus region specifically includes:

[0066] Any two adjacent elements in the eight-chain code sequence are taken as an element group, and the absolute value of the difference between the two elements in each element group is taken as the element difference value of each element group. The discrete degrees of the element difference values ​​of all element groups are analyzed and negative correlation mapping is performed to obtain the non-circular degree of the cell nucleus area.

[0067] In the embodiment of the present invention, the analysis of the discreteness of the element difference values ​​of all element groups may be achieved by calculating the standard deviation or variance of the element difference values ​​of all element groups, which is not limited here.

[0068] As an example, in one embodiment of the present invention, the expression of the non-circularity degree of the cell nucleus region can be specifically, for example, as follows:

[0069] ;

[0070] in, Indicates the non-circularity of the nuclear region; Indicates The difference value of the elements of the element group; represents the average of the element difference values ​​of all element groups; Indicates the number of element groups; represents the variance of the element difference values ​​of all element groups; Indicates the preset first adjustment parameter, which is used to prevent the denominator from being 0. The value range is In one embodiment of the present invention, Set to 0.001, The specific value of can also be set by the implementer according to the specific implementation scenario and is not limited here.

[0071] It should be noted that in other embodiments of the present invention, the positive and negative correlations may be represented by other basic mathematical operations, which will not be elaborated herein.

[0072] Step S202: Obtain the grayscale distribution characteristic value and the number distribution characteristic value of the stem cell region, the grayscale distribution characteristic value is positively correlated with the average grayscale value of all pixels in the cytoplasm region, the grayscale distribution characteristic value is negatively correlated with the average grayscale value of all pixels in the nucleus region, the number distribution characteristic value is positively correlated with the number of all pixels in the cytoplasm region, and the number distribution characteristic value is negatively correlated with the number of all pixels in the nucleus region.

[0073] From the above analysis, it can be seen that as the differentiation of stem cells proceeds, the grayscale value and number of pixels in the cytoplasm region will gradually increase, while the grayscale value and number of pixels in the nucleus region will gradually decrease. Therefore, the grayscale distribution characteristic value and the number distribution characteristic value of the stem cell region can be obtained, wherein the grayscale distribution characteristic value is positively correlated with the average grayscale value of all pixels in the cytoplasm region, the grayscale distribution characteristic value is negatively correlated with the average grayscale value of all pixels in the nucleus region, the number distribution characteristic value is positively correlated with the number of all pixels in the cytoplasm region, and the number distribution characteristic value is negatively correlated with the number of all pixels in the nucleus region. The larger the grayscale distribution characteristic value, the larger the grayscale value of the pixels in the cytoplasm region, and the smaller the grayscale value of the pixels in the nucleus region. The larger the number distribution characteristic value, the more pixels in the cytoplasm region, and the fewer pixels in the nucleus region, which further indicates that the degree of differentiation of the stem cells to be tested is higher.

[0074] As an example, in one embodiment of the present invention, the expressions of the grayscale distribution characteristic value and the quantity distribution characteristic value of the stem cell region can be specifically, for example, as follows:

[0075] ;

[0076] ;

[0077] in, Represents the grayscale distribution characteristic value of the stem cell area; Indicates the quantitative distribution characteristic value of the stem cell area; Represents the average gray value of all pixels in the cytoplasm area; Represents the average gray value of all pixels in the cell nucleus area; Represents the number of all pixels in the cytoplasm area; Represents the number of all pixels in the cell nucleus area; and They represent the preset second adjustment parameter and the preset third adjustment parameter respectively, to prevent the denominator from being 0. and The value range of In one embodiment of the present invention, and Set to 0.001, and The specific value of can also be set by the implementer according to the specific implementation scenario and is not limited here.

[0078] Step S203: negative correlation mapping is performed on the number of edge lines in the stem cell region to obtain a characteristic value of the number of organelles in the stem cell region.

[0079] As stem cells differentiate into red blood cells, they gradually lose their organelles. That is to say, as the degree of differentiation deepens, the number of organelles in stem cells will become fewer and fewer. The edge lines in the stem cell area can be used to characterize the edges of organelles. Therefore, the number of edge lines in the stem cell area can be negatively correlated to obtain the characteristic value of the number of organelles in the stem cell area. The larger the characteristic value of the number of organelles, the fewer the number of organelles in the tested stem cells, and thus the higher the degree of differentiation of the tested stem cells.

[0080] As an example, in one embodiment of the present invention, the expression of the characteristic value of the number of organelles in the stem cell region can be specifically, for example, as follows:

[0081] ;

[0082] in, The characteristic value of the number of organelles representing the stem cell area; represents the number of margin lines in the stem cell area; Indicates the preset fourth adjustment parameter, which is used to prevent the denominator from being 0. The value range is In one embodiment of the present invention, Set to 0.001, The specific value of can also be set by the implementer according to the specific implementation scenario and is not limited here.

[0083] Step S204: The non-circularity degree, grayscale distribution characteristic value, quantity distribution characteristic value and organelle quantity characteristic value are integrated and normalized to obtain the structural differentiation factor of the stem cell to be tested.

[0084] Through the above analysis, it can be seen that the larger the non-circularity degree, grayscale distribution characteristic value, number distribution characteristic value and organelle number characteristic value are, the greater the degree of differentiation of the stem cell to be tested is. Therefore, the non-circularity degree, grayscale distribution characteristic value, number distribution characteristic value and organelle number characteristic value can be integrated and normalized to obtain the structural differentiation factor of the stem cell to be tested.

[0085] In the embodiment of the present invention, the synthesis can be achieved by calculating the sum or product of the non-circularity degree, the grayscale distribution characteristic value, the number distribution characteristic value and the organelle number characteristic value, which is not limited here.

[0086] As an example, in one embodiment of the present invention, the expression of the structural differentiation factor of the stem cell to be tested can be specifically, for example, as follows:

[0087] ;

[0088] in, represents the structural differentiation factor of the stem cell to be tested; Indicates the non-circularity of the nuclear region; Represents the grayscale distribution characteristic value of the stem cell area; Indicates the quantitative distribution characteristic value of the stem cell area; The characteristic value of the number of organelles representing the stem cell area; Represents the hyperbolic tangent function, which is used for normalization.

[0089] Step S3: perform ellipse fitting on the edge of the stem cell region to obtain the ellipse-like coefficient of the stem cell region; obtain the morphological differentiation factor of the stem cell to be tested based on the change in the gray value of the pixel point between each edge pixel point and the centroid of the stem cell region and the ellipse-like coefficient; obtain the differentiation evaluation value of the stem cell to be tested based on the structural differentiation factor and the morphological differentiation factor.

[0090] In the process of stem cell differentiation into red blood cells, its overall shape will gradually change into a disc shape, and a concave phenomenon will appear on its inner side. In the image, the edge shape of the stem cell area is close to an ellipse, and the change of the gray value of the pixel point between the edge pixel point and the centroid of the stem cell area is different. At the same time, the gray value change difference between the inner and outer sides of the stem cell area is large. Therefore, the edge of the stem cell area can be firstly fitted with an ellipse to obtain the ellipse-like coefficient of the stem cell area, which provides a data basis for the subsequent calculation of the morphological differentiation factor of the stem cell to be tested.

[0091] Preferably, in one embodiment of the present invention, the method for obtaining the ellipse-like coefficient of the stem cell region specifically includes:

[0092] First, the least square method is used to perform ellipse fitting on the edge of the stem cell region to obtain the fitting loss value of the stem cell region and the major axis and minor axis of the fitting ellipse. In other embodiments of the present invention, the maximum likelihood method may also be used to implement ellipse fitting, which is not limited here.

[0093] Since the overall morphology of the tested stem cells will develop towards a relatively flat disc shape during the differentiation process into red blood cells, as the differentiation deepens, the shape of the edge of the stem cell region in the image tends to be more like a flat ellipse, and the more pixels there are on the major axis of the fitted ellipse, the longer the major axis is, and the fewer pixels there are on the minor axis, the shorter the minor axis is, and the greater the difference between the lengths of the major axis and the minor axis, the flatter the fitted ellipse is. Therefore, the flatness coefficient of the stem cell region can be obtained according to the difference between the number of pixels on the major axis and the number of pixels on the minor axis of the fitted ellipse.

[0094] In one embodiment of the present invention, the difference between the number of pixel points on the major axis and the number of pixel points on the minor axis of the fitted ellipse may be used as a flattening coefficient to analyze the difference between the two.

[0095] The smaller the fitting loss value is, the closer the edge shape of the stem cell region is to an ellipse. Therefore, the fitting loss value can be negatively correlated to obtain the ellipse similarity of the stem cell region.

[0096] Then, the flatness coefficient and the ellipse similarity are integrated to obtain the ellipse-like coefficient of the stem cell region. In an embodiment of the present invention, the sum or product of the flatness coefficient and the ellipse similarity can be used as the ellipse-like coefficient of the stem cell region to achieve the integration of the two, which is not limited here.

[0097] As an example, in one embodiment of the present invention, the expression of the elliptic coefficient of the stem cell region can be specifically, for example, as follows:

[0098] ;

[0099] in, represents the ellipse-like coefficient of the stem cell area; Indicates the number of pixels on the major axis of the fitted ellipse; Indicates the number of pixels on the minor axis of the fitted ellipse; represents the fitting loss value of the stem cell region; represents the flatness coefficient of the stem cell area; Similarity of ellipses representing stem cell regions; Indicates the preset fifth adjustment parameter, which is used to prevent the denominator from being 0. The value range is In one embodiment of the present invention, Set to 0.001, The specific value of can also be set by the implementer according to the specific implementation scenario and is not limited here.

[0100] At the same time, the disc shape of the stem cells to be tested during the differentiation process will also lead to different changes in the grayscale values ​​of the pixels between the edge pixels and the centroid of the stem cell area in the image, and its morphology is manifested as an inner concave phenomenon, which will lead to a large difference in the grayscale value changes between the inner and outer sides of the stem cell area. Therefore, the changes in the grayscale values ​​of the pixels between each edge pixel and the centroid of the stem cell area can be further analyzed. At the same time, combined with the ellipse-like coefficient of the stem cell area, the morphological differentiation factor obtained can reflect the overall morphological characteristics of the stem cells as they differentiate. The larger the morphological differentiation factor, the greater the degree of differentiation of the stem cells to be tested. Subsequently, the degree of differentiation of the stem cells to be tested can be accurately identified and evaluated in combination with the morphological differentiation factor and the structural differentiation factor. Among them, the acquisition of the centroid is a technical means well known to those skilled in the art and will not be elaborated here.

[0101] Preferably, in one embodiment of the present invention, the method for obtaining the structural differentiation factor of the stem cell to be tested specifically includes:

[0102] See also Figure 3 , which shows a flow chart of a method for obtaining morphological differentiation factors of stem cells to be tested provided by one embodiment of the present invention.

[0103] Step S301: taking the line between each edge pixel point and the centroid of the stem cell region as a reference line segment, and obtaining the first morphological feature value of the stem cell region according to the difference in grayscale value changes of the pixels on each reference line segment and the ellipse-like coefficient.

[0104] First, the line between each edge pixel point and the centroid of the stem cell region is used as a reference line segment. As the differentiation of the tested stem cells deepens, the flat disc shape they present becomes more obvious, and the difference in the change of the grayscale value of the pixels on each reference line segment of the stem cell region that appears as an ellipse in the image becomes larger. Therefore, the difference in the change of the grayscale value of the pixels on each reference line segment can be analyzed, and the first morphological characteristic value of the stem cell region can be obtained by combining the ellipse-like coefficient of the stem cell region. The larger the first morphological characteristic value, the closer the edge shape of the stem cell region is to a flat ellipse, and the larger the difference in the grayscale value change between the edge pixel point and the centroid of the stem cell region is, which further indicates that the morphology of the tested stem cells is closer to a flat disc.

[0105] Preferably, in one embodiment of the present invention, the method for acquiring the first morphological characteristic value of the stem cell region specifically includes:

[0106] Take any reference line segment as the target reference line segment, take the absolute value of the difference between the grayscale value of each pixel point on the target reference line segment and the next adjacent pixel point as the grayscale change value of each pixel point, analyze the overall level of the grayscale change values ​​of all pixels on the target reference line segment, and obtain the grayscale change characteristic value of the target reference line segment. The grayscale change characteristic value reflects the change of the grayscale value of the pixel point on the target reference line segment.

[0107] In an embodiment of the present invention, the average or median of the grayscale change values ​​of all pixels on the target reference line segment can be used as the grayscale change characteristic value of the target reference line segment to achieve an overall level analysis of the grayscale change values ​​of all pixels on the target reference line segment, which is not limited here.

[0108] It should be noted that, for the existing boundary problem, if a pixel point has no next adjacent pixel point, the grayscale value of the pixel point is used as the grayscale change value of the pixel point to solve the existing boundary problem and facilitate the smooth progress of subsequent calculations.

[0109] Take any two reference line segments as a reference line segment group, and take the absolute value of the difference between the grayscale change characteristic values ​​of the two reference line segments in each reference line segment group as the initial grayscale change difference of each reference line segment group; the smaller the angle between the two reference line segments in the reference line segment group, the greater the confidence in the analysis of the grayscale value change difference of the pixels on the two reference line segments. Therefore, the angle between the two reference line segments in each reference line segment group is negatively correlated and normalized to obtain the adjustment weight of each reference line segment group, and then use the adjustment weight to weight the initial grayscale change difference, so that the difference in the grayscale value change characteristics of the pixels on the two reference line segments in the reference line segment group is accurately reflected by the obtained real grayscale change difference.

[0110] In one embodiment of the present invention, the natural constant A negative exponential function with base The normalization processing of negative correlation is realized in the form of a function, which is not limited here.

[0111] It should be noted that the two reference line segments in each reference line segment group can form two included angles. Therefore, in the embodiment of the present invention, the value range of the included angle is set to .

[0112] The actual grayscale change differences of all reference line segment groups are integrated to obtain the comprehensive grayscale change differences of the stem cell region. The larger the comprehensive grayscale change difference is, the more likely the surface morphology of the tested stem cell is to be disc-shaped, which further indicates that the degree of differentiation of the tested stem cell is greater.

[0113] In the embodiment of the present invention, the sum or product of the real grayscale change differences of all reference line segment groups can be used as the comprehensive grayscale change difference of the stem cell region to achieve the integration of the real grayscale change differences of all reference line segment groups, which is not limited here.

[0114] Then, the comprehensive grayscale change difference and the ellipse-like coefficient are integrated and normalized to obtain the first morphological characteristic value of the stem cell area.

[0115] In the embodiment of the present invention, the sum or product of the comprehensive grayscale change difference and the ellipse-like coefficient can be used as the first morphological characteristic value of the stem cell region to achieve the integration of the two, which is not limited here.

[0116] As an example, in one embodiment of the present invention, the expression of the first morphological characteristic value of the stem cell region may be specifically, for example, as follows:

[0117] ;

[0118] ;

[0119] in, represents the first morphological characteristic value of the stem cell area; represents the ellipse-like coefficient of the stem cell area; Indicates the difference in comprehensive grayscale changes in the stem cell area; Indicates The angle between two reference line segments in a reference line segment group; Indicates The adjustment weights of the reference line segment groups; and Respectively represent Grayscale change characteristic values ​​of two reference line segments in a reference line segment group; Indicates The difference in initial grayscale changes of the reference line segment groups; Indicates the number of reference line segment groups; Indicates The actual grayscale change difference of the reference line segment groups; Indicated by natural constant An exponential function with base ; Represents the hyperbolic tangent function, which is used for normalization.

[0120] Step S302: The pixel point with the largest grayscale value on each reference line segment is used as the grayscale turning point of each reference line segment, the line segment between the centroid and the grayscale turning point is used as the inner line segment of each reference line segment, and the line segment between the edge pixel point and the grayscale turning point is used as the outer line segment of each reference line segment.

[0121] As the stem cells to be tested differentiate into red blood cells, depressions will appear on the inner side of their surface, and the greater the degree of differentiation, the more obvious the depression, resulting in specific characteristics in the grayscale values ​​of the pixels on the inner and outer sides of the stem cell area. Specifically, the grayscale value of the pixels from the centroid of the stem cell area to the junction of the depression gradually increases, and the grayscale value of the pixels from the edge of the stem cell area to the junction of the depression also gradually increases, and the grayscale value of the pixels from the centroid to the junction of the depression changes faster. Therefore, the pixel with the largest grayscale value on each reference line segment is first taken as the grayscale turning point of each reference line segment, and the grayscale turning point is the pixel at the junction of the depression. The line segment between the centroid and the grayscale turning point is taken as the inner line segment of each reference line segment, and the line segment between the edge pixel and the grayscale turning point is taken as the outer line segment of each reference line segment. The pixels on the inner line segment represent the pixels on the inner side of the stem cell area, and the pixels on the outer line segment represent the pixels on the outer side of the stem cell area.

[0122] Step S303: obtaining a second morphological characteristic value of the stem cell region according to the number of pixels on the inner and outer segments of each reference line segment and the change in the grayscale value of the pixels on the inner segment.

[0123] The more pixels there are on the inner segment of the reference line segment than on the outer segment, and the faster the grayscale value of the pixels on the inner segment changes, the more obvious the inner depression of the stem cell to be tested is, and thus the greater the degree of differentiation of the stem cell to be tested is. Therefore, the number of pixels on the inner and outer segments of each reference line segment, as well as the change in the grayscale value of the pixels on the inner segment can be analyzed to obtain the second morphological characteristic value of the stem cell area. The larger the second morphological characteristic value is, the more obvious the inner depression of the stem cell to be tested is.

[0124] Preferably, in one embodiment of the present invention, the method for acquiring the second morphological characteristic value of the stem cell region specifically includes:

[0125] The overall level of grayscale change values ​​of all pixels on the inner line segment of each reference line segment is analyzed to obtain a first concavity parameter of each reference line segment.

[0126] In an embodiment of the present invention, the average value or median of the grayscale change values ​​of all pixels on the inner line segment of each reference line segment can be used as the first concave parameter of each reference line segment to realize the overall level analysis of the grayscale change values ​​of all pixels on the inner line segment of each reference line segment, which is not limited here.

[0127] The number of pixels on the inner line segment of each reference line segment is used as the numerator, the number of pixels on the outer line segment of each reference line segment is used as the denominator, and the ratio is used as the second concave parameter of each reference line segment;

[0128] The first concave parameter and the second concave parameter are integrated to obtain a comprehensive concave parameter of each reference line segment; the overall level of the comprehensive concave parameters of all reference line segments is analyzed and normalized to obtain a second morphological characteristic value of the stem cell region.

[0129] In the embodiment of the present invention, the sum or product of the first concave parameter and the second concave parameter of each reference line segment may be used as the comprehensive concave parameter of each reference line segment to achieve the combination of the two, which is not limited here.

[0130] In the embodiment of the present invention, the overall level analysis of the comprehensive concavity parameters of all reference line segments may be achieved by calculating the average value or median of the comprehensive concavity parameters of all reference line segments, which is not limited here.

[0131] As an example, in one embodiment of the present invention, the expression of the second morphological characteristic value of the stem cell region may be specifically, for example, as follows:

[0132] ;

[0133] in, The second morphological characteristic value representing the stem cell area; Indicates The number of pixels on the inner line segment of the reference line segment; Indicates The number of pixels on the outer line segment of the reference line segment; Indicates The second concavity parameter of the reference line segment is not equal to 0, because the stem cells to be tested do not have a concavity on the entire surface, but only on the inner part. ; Indicates The first concavity parameter of a reference line segment; Indicates The comprehensive concavity parameters of the reference line segments; Indicates the number of reference line segments; Represents the hyperbolic tangent function, which is used for normalization.

[0134] Step S304: The first morphological characteristic value and the second morphological characteristic value are integrated and normalized to obtain the morphological differentiation factor of the stem cell to be tested.

[0135] The larger the first morphological characteristic value and the second morphological characteristic value are, the more the morphology of the stem cell to be tested is consistent with the morphology of mature red blood cells, and further the greater the degree of differentiation of the stem cell to be tested is. Therefore, the first morphological characteristic value and the second morphological characteristic value can be integrated and normalized to obtain the morphological differentiation factor of the stem cell to be tested.

[0136] In the embodiment of the present invention, the sum or product of the first morphological characteristic value and the second morphological characteristic value may be used as the morphological differentiation factor of the stem cell to be tested to achieve a combination of the two, which is not limited here.

[0137] As an example, in one embodiment of the present invention, the expression of the morphological differentiation factor of the stem cell to be tested can be specifically, for example, as follows:

[0138] ;

[0139] in, represents the morphological differentiation factor of the stem cells to be tested; represents the first morphological characteristic value of the stem cell area; The second morphological characteristic value representing the stem cell area; Represents the hyperbolic tangent function, which is used for normalization.

[0140] From the above analysis, it can be seen that the larger the structural differentiation factor and morphological differentiation factor of the stem cell to be tested, the greater the degree of differentiation of the stem cell to be tested. Therefore, based on the structural differentiation factor and the morphological differentiation factor, the differentiation evaluation value of the stem cell to be tested can be obtained, and the degree of differentiation of the stem cell to be tested can be accurately identified through the differentiation evaluation value.

[0141] Preferably, in one embodiment of the present invention, the method for obtaining the differentiation evaluation value of the stem cell to be tested specifically includes:

[0142] The sum of the structural differentiation factor and the morphological differentiation factor is normalized to obtain the differentiation evaluation value of the stem cell to be tested.

[0143] In other embodiments of the present invention, the product value of the structural differentiation factor and the morphological differentiation factor may also be normalized to obtain the differentiation evaluation value of the stem cell to be tested, which is not limited here.

[0144] As an example, in one embodiment of the present invention, the expression of the differentiation evaluation value of the stem cell to be tested can be specifically, for example, as follows:

[0145] ;

[0146] in, represents the differentiation evaluation value of the stem cell to be tested; represents the structural differentiation factor of the stem cell to be tested; represents the morphological differentiation factor of the stem cells to be tested; Represents the hyperbolic tangent function, which is used for normalization.

[0147] Step S4: Based on the differentiation evaluation value, the differentiation degree of the stem cells to be tested is evaluated.

[0148] The larger the differentiation evaluation value is, the greater the degree of differentiation of the tested stem cell into red blood cells is. Therefore, the degree of differentiation of the tested stem cell can be accurately evaluated based on the differentiation evaluation value.

[0149] Preferably, in one embodiment of the present invention, the method for evaluating the differentiation degree of the stem cells to be tested specifically includes:

[0150] If the differentiation evaluation value is less than the preset first threshold, it means that the stem cell to be tested is in the early stage of differentiation, and the degree of differentiation of the stem cell to be tested is considered to be low. If the differentiation evaluation value is not less than the preset first threshold and less than the preset second threshold, it means that the stem cell to be tested is in the middle stage of differentiation, and the degree of differentiation of the stem cell to be tested is considered to be medium. If the differentiation evaluation value is not less than the preset second threshold, it means that the stem cell to be tested is in the late stage of differentiation, and the degree of differentiation of the stem cell to be tested is considered to be high. The value range of the preset first threshold is , the value range of the preset second threshold is In one embodiment of the present invention, the preset first threshold and the preset second threshold are set to 0.3 and 0.7 respectively. The specific values ​​of the preset first threshold and the preset second threshold can also be set by the implementer according to the specific implementation scenario, which is not limited here.

[0151] In summary, the embodiment of the present invention first obtains a grayscale image of the stem cell to be tested; performs edge detection on the grayscale image to obtain a stem cell region and edge lines in the stem cell region; performs threshold segmentation on the stem cell region to obtain a cytoplasm region and a nucleus region; obtains a structural differentiation factor of the stem cell to be tested based on the grayscale value and number of pixels in the cytoplasm region and the nucleus region, the position distribution of edge pixels in the nucleus region, and the number of edge lines in the stem cell region; performs ellipse fitting on the edge of the stem cell region to obtain a quasi-ellipse coefficient of the stem cell region; obtains a morphological differentiation factor of the stem cell to be tested based on the change in the grayscale value of the pixel point between each edge pixel point and the centroid of the stem cell region and the quasi-ellipse coefficient; obtains a differentiation evaluation value of the stem cell to be tested based on the structural differentiation factor and the morphological differentiation factor; and evaluates the degree of differentiation of the stem cell to be tested based on the differentiation evaluation value.

[0152] It should be noted that the sequence of the above embodiments of the present invention is for description only and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0153] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for identifying the degree of stem cell differentiation based on image features, characterized in that: The method comprises: Acquire a grayscale image of the stem cells to be tested; perform edge detection on the grayscale image to obtain a stem cell region and an edge line in the stem cell region; Performing threshold segmentation on the stem cell region to obtain a cytoplasm region and a cell nucleus region; obtaining a structural differentiation factor of the stem cell to be tested according to the grayscale value and number of pixel points in the cytoplasm region and the cell nucleus region, the position distribution of edge pixel points in the cell nucleus region, and the number of edge lines in the stem cell region; Performing ellipse fitting on the edge of the stem cell region to obtain a quasi-ellipse coefficient of the stem cell region; obtaining a morphological differentiation factor of the stem cell to be tested based on the change in the gray value of the pixel point between each edge pixel point and the centroid of the stem cell region and the quasi-ellipse coefficient; obtaining a differentiation evaluation value of the stem cell to be tested based on the structural differentiation factor and the morphological differentiation factor; Based on the differentiation evaluation value, evaluating the differentiation degree of the stem cell to be tested; The method of obtaining the structural differentiation factor of the stem cell to be tested comprises: Obtaining an eight-chain code sequence of edge pixels of a cell nucleus region, and obtaining a non-circular degree of the cell nucleus region according to differences between adjacent elements in the eight-chain code sequence; Obtaining a grayscale distribution characteristic value and a quantity distribution characteristic value of the stem cell region, wherein the grayscale distribution characteristic value is positively correlated with an average grayscale value of all pixels in the cytoplasm region, the grayscale distribution characteristic value is negatively correlated with an average grayscale value of all pixels in the cell nucleus region, the quantity distribution characteristic value is positively correlated with the number of all pixels in the cytoplasm region, and the quantity distribution characteristic value is negatively correlated with the number of all pixels in the cell nucleus region; Negative correlation mapping is performed on the number of edge lines in the stem cell area to obtain the characteristic value of the number of organelles in the stem cell area; The non-circularity degree, the grayscale distribution characteristic value, the quantity distribution characteristic value and the organelle quantity characteristic value are integrated and normalized to obtain a structural differentiation factor of the stem cell to be tested; The morphological differentiation factors of the stem cells to be tested are obtained by: The line between each edge pixel point and the centroid of the stem cell region is used as a reference line segment; Obtaining a first morphological characteristic value of the stem cell region according to the difference in grayscale value changes of the pixels on each reference line segment and the ellipse-like coefficient; The pixel point with the maximum gray value on each reference line segment is taken as the gray turning point of each reference line segment, the line segment between the centroid and the gray turning point is taken as the inner line segment of each reference line segment, and the line segment between the edge pixel point and the gray turning point is taken as the outer line segment of each reference line segment; Obtaining a second morphological characteristic value of the stem cell region according to the number of pixels on the inner and outer segments of each reference line segment and the change in the grayscale value of the pixels on the inner segment; The first morphological characteristic value and the second morphological characteristic value are integrated and normalized to obtain a morphological differentiation factor of the stem cell to be tested; The step of obtaining the differentiation evaluation value of the stem cell to be tested based on the structural differentiation factor and the morphological differentiation factor comprises: The sum of the structural differentiation factor and the morphological differentiation factor is normalized to obtain a differentiation evaluation value of the stem cell to be tested.

2. The method for identifying the degree of stem cell differentiation based on image features according to claim 1, characterized in that: The step of obtaining the non-circularity degree of the cell nucleus region according to the difference between adjacent elements in the eight-chain code sequence comprises: Any two adjacent elements in the eight-chain code sequence are regarded as an element group, and the absolute value of the difference between the two elements in each element group is regarded as the element difference value of each element group; After analyzing the discrete degrees of the element difference values ​​of all element groups and performing negative correlation mapping, the non-circular degree of the cell nucleus region is obtained.

3. The method for identifying the degree of stem cell differentiation based on image features according to claim 1, characterized in that: The performing ellipse fitting on the edge of the stem cell region to obtain the ellipse-like coefficient of the stem cell region comprises: Using the least square method, an ellipse is fitted to the edge of the stem cell region to obtain the fitting loss value of the stem cell region and the major axis and minor axis of the fitting ellipse; The flatness coefficient of the stem cell region is obtained according to the difference between the number of pixels on the major axis and the number of pixels on the minor axis of the fitted ellipse; Performing negative correlation mapping on the fitting loss value to obtain the ellipse similarity of the stem cell region; The flatness coefficient and the ellipse similarity are integrated to obtain the ellipse-like coefficient of the stem cell region.

4. The method for identifying the degree of stem cell differentiation based on image features according to claim 1, characterized in that: The obtaining of the first morphological characteristic value of the stem cell region comprises: Taking any reference line segment as the target reference line segment, taking the absolute value of the difference between the grayscale value of each pixel point on the target reference line segment and the next adjacent pixel point as the grayscale change value of each pixel point, analyzing the overall level of the grayscale change values ​​of all pixels on the target reference line segment, and obtaining the grayscale change characteristic value of the target reference line segment; Take any two reference line segments as reference line segment groups, and take the absolute value of the difference between the grayscale change feature values ​​of the two reference line segments in each reference line segment group as the initial grayscale change difference of each reference line segment group; perform negative correlation normalization processing on the angle between the two reference line segments in each reference line segment group to obtain the adjustment weight of each reference line segment group, and use the adjustment weight to weight the initial grayscale change difference to obtain the real grayscale change difference of each reference line segment group, where the value range of the angle is ; The real grayscale change differences of all reference line segment groups are integrated to obtain the integrated grayscale change differences of the stem cell region; The comprehensive grayscale change difference and the ellipse-like coefficient are integrated and normalized to obtain a first morphological characteristic value of the stem cell region.

5. The method for identifying the degree of differentiation of stem cells based on image features according to claim 4, characterized in that: The obtaining of the second morphological characteristic value of the stem cell region comprises: Analyze the overall level of the grayscale change values ​​of all pixels on the inner line segment of each reference line segment to obtain a first concave parameter of each reference line segment; The number of pixels on the inner line segment of each reference line segment is used as the numerator, the number of pixels on the outer line segment of each reference line segment is used as the denominator, and the ratio is used as the second concave parameter of each reference line segment; The first concave parameter and the second concave parameter are integrated to obtain a comprehensive concave parameter of each reference line segment; the overall level of the comprehensive concave parameters of all reference line segments is analyzed and normalized to obtain a second morphological characteristic value of the stem cell region.

6. The method for identifying the degree of differentiation of stem cells based on image features according to claim 1, characterized in that: The step of evaluating the differentiation degree of the stem cell to be tested based on the differentiation evaluation value comprises: If the differentiation evaluation value is less than the preset first threshold, the degree of differentiation of the stem cells to be tested is low; if the differentiation evaluation value is not less than the preset first threshold and is less than the preset second threshold, the degree of differentiation of the stem cells to be tested is medium; if the differentiation evaluation value is not less than the preset second threshold, the degree of differentiation of the stem cells to be tested is high.

7. The method for identifying the degree of differentiation of stem cells based on image features according to claim 6, characterized in that: The value range of the preset first threshold is , the value range of the preset second threshold is .