A platelet microtubule classification method and system based on super-resolution images

Abstract

The application discloses a kind of based on super-resolution image's platelet microtubule classification method and classification system, the classification method, including the following steps: obtaining immunofluorescence labeled platelet microtubule super-resolution image;With single platelet microtubule structure distribution as classification object, according to classification principle is divided into "regular distribution", "diffuse distribution", "aggregated distribution" and "irregular distribution" four categories, the classification method can be used to classify platelet microtubule in sample, and it has important significance to platelet physiological function research.In addition, based on the classification method, the platelet microtubule classification system based on super-resolution image provided by the application can automatically extract the complete image of single platelet microtubule based on the obtained platelet microtubule super-resolution image, and mark the type of microtubule according to the classification principle, and automatically display the classification result, which can automatically classify platelet microtubule based on platelet microtubule super-resolution image.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of platelet microtubule classification, and more particularly relates to a platelet microtubule classification method and system based on super-resolution images. BACKGROUND

[0002] Platelets are anucleate cell fragments with a diameter of about 2-4 μm and a thickness of 0.2-1.5 μm, which participate in a series of physiological and pathological processes in the human body, including maintaining the integrity of blood vessels, promoting tumor growth and migration, etc. Under the plasma membrane of platelets, 3-24 microtubules are parallel to the plasma membrane, which together form a ring-shaped marginal band to maintain the discoid shape of resting platelets. Microtubules are a dynamic structure that is sensitive to physiological and pathological conditions: when platelets are exposed to a 4-degree Celsius environment, microtubules are prone to depolymerization, resulting in the loss of discoid shape of platelets and the existence of spherical shape; in the early activation process of platelets, microtubules slide, extend, and the microtubule band is entangled, realizing the transformation of platelet discoid shape to spherical shape and the centralization of granules; in addition, microtubules also change under pathological conditions. Therefore, studying how to classify single platelet microtubules has important significance for the research and detection of physiological functions of platelets. However, the research is still in its initial stage, and there is still a lack of classification and analysis of different morphological microtubules.

[0003] The classification of platelet microtubules is based on the fine detection of microtubule structure, but ordinary microscopic detection technology is limited by the diffraction limit, and it is difficult to distinguish the morphology of microtubules in a single platelet. Although electron microscopy can observe the fine structure of microtubules, such as the number, length, and morphology of microtubules, electron microscopy has the disadvantages of small imaging range, low efficiency, long time consumption, high cost, complex operation, and the need for professional identification, which makes it difficult to perform classification analysis and cannot meet the demand for rapid detection of a large number of blood samples in clinical practice. At present, the classification of microtubules is insufficient, and further development of technology for platelet microtubule classification is needed. SUMMARY

[0004] In view of the above defects or improvement needs of the prior art, the present application provides a platelet microtubule classification method and system based on super-resolution images, which aims to extract labeled single platelet microtubule structures as classification objects based on platelet super-resolution images, and classify them into "regular distribution", "diffuse distribution", "aggregated distribution", and "irregular distribution" according to different microtubule structures, thereby solving the technical problem of the lack of methods and systems for classifying microtubule structures in the prior art.

[0005] To achieve the above-mentioned purpose, according to one aspect of the present application, a platelet microtubule classification method based on super-resolution images is provided, which comprises the following steps:

[0006] obtaining a single platelet microtubule image by super-resolution imaging of microtubule immunofluorescence labeling of a single platelet and binarizing the single platelet microtubule image to obtain a single platelet microtubule image composed of fluorescent signal pixels and non-fluorescent signal pixels;

[0007] for the single platelet microtubule image, using an image segmentation algorithm to segment the fluorescent signal pixel region into independent microtubule pixel regions to obtain a single platelet microtubule distribution image;

[0008] Classify the distribution form of the microtubules involved in the single platelet microtubule distribution image, specifically as follows:

[0009] (1) Search for an ellipse in the single platelet microtubule image that satisfies condition A and has the smallest area, the ratio of the major axis to the minor axis is in [0.2, 1], and the ellipse area threshold is in (0.8 μm 2 -30 μm 2 ],

[0010] If the ratio of the area occupied by the single platelet microtubule fluorescent signal to the area of the ellipse is in a preset range 1, it is determined that the microtubules involved in the single platelet microtubule image are "regularly distributed", and the preset range 1 is in (0, 0.6);

[0011] If the ratio of the area occupied by the single platelet microtubule fluorescent signal to the area of the ellipse is in a preset range 2, it is determined that the microtubules involved in the single platelet microtubule image are "diffusely distributed", and the preset range 2 is in [0.6, 1];

[0012] (2) Search for an ellipse in the single platelet microtubule image that satisfies condition A and has the smallest area, the ratio of the major axis to the minor axis is in [0.2, 1], and the ellipse area threshold is in (0.1 μm 2 , 0.8 μm 2 ], and the ratio of the area occupied by the single platelet microtubule fluorescent signal to the area of the ellipse is in a preset range 3, it is determined that the microtubules involved in the single platelet microtubule image are "aggregated distribution", and the preset range 3 is in [0.6, 1];

[0013] (3) If the single platelet microtubule image does not belong to "regular distribution", "diffuse distribution" and "aggregated distribution", it is determined that the microtubules involved in the single platelet microtubule image are "irregular distribution";

[0014] Condition A: Microtubules exceeding the preset ellipse proportion threshold are distributed inside the ellipse; the preset ellipse proportion threshold is above 95%.

[0015] Preferably, the platelet microtubule classification method based on super-resolution images, for the microtubules involved in the microtubule image within the ellipse, the ratio of the outer edge length of the microtubule image within the ellipse to the circumference of the ellipse is in the preset range, and the subclass classification is as follows:

[0016] If the ratio of the outer edge length of the microtubule image within the ellipse to the circumference of the ellipse is in (0, 0.2), it is a "cross coil";

[0017] If the ratio of the outer edge length of the microtubule image within the ellipse to the circumference of the ellipse is in (0.75, 1], it is a "ring distribution".

[0018] Preferably, the platelet microtubule classification method based on super-resolution images, the ratio is calculated according to the following formula:

[0019] E = L1 / L2

[0020] In the formula, L1 is the outer edge length of the microtubule image within the ellipse, and L2 is the circumference of the ellipse.

[0021] Preferably, the platelet microtubule classification method based on super-resolution images, the ratio of the long axis to the short axis of the ellipse is in [0.5, 1], and the area threshold of the ellipse is in [1.7 μm 2 -12.6 μm 2 ].

[0022] Preferably, the platelet microtubule classification method based on super-resolution images, the ratio of the long axis to the short axis of the ellipse is in [0.5, 1], and the area threshold of the ellipse is in [0.2 μm 2 , 0.8 μm 2 ].

[0023] According to another aspect of the present application, a platelet microtubule classification system based on super-resolution images is provided, which comprises a platelet microtubule super-resolution image acquisition module, a single platelet microtubule image extraction and labeling module, and a classification module.

[0024] The platelet microtubule super-resolution image acquisition module is used to obtain a single platelet microtubule image composed of fluorescent signal pixels and non-fluorescent signal pixels by binarization, and submit it to the single platelet microtubule distribution image extraction and labeling module;

[0025] The single platelet microtubule image extraction and labeling module is used to divide the fluorescent signal pixel area in the single platelet microtubule image into independent microtubule pixel areas by using an image segmentation algorithm, to obtain a single platelet microtubule distribution image, and submit it to the classification module;

[0026] The classification module is used to classify the platelets according to the distribution pattern of microtubules, as follows:

[0027] The microtubules are classified according to their distribution patterns in the images of microtubule distribution within a single platelet, as follows:

[0028] (1) Search for the ellipse with the smallest area in the microtubule image of a single platelet that satisfies condition A, wherein the ratio of the major axis to the minor axis of the ellipse is in the range [0.2, 1], and the ellipse area threshold is in the range [0.8 μm]. 2 -30μm 2 ],

[0029] If the ratio of the area occupied by the fluorescence signal of microtubules in a single platelet to the area of ​​the ellipse is within a preset range of 1, then the microtubules involved in the microtubule image of a single platelet are determined to be "regularly distributed", and the preset range of 1 is (0, 0.6).

[0030] If the ratio of the area occupied by the fluorescence signal of the microtubules in a single platelet to the area of ​​the ellipse is within a preset range of 2, then the microtubules involved in the microtubule image of a single platelet are determined to be "diffusely distributed", and the preset range of 2 is in [0.6, 1].

[0031] (2) Search for the ellipse in the microtubule image of the single platelet that satisfies condition A and has the smallest area, wherein the ratio of the major axis to the minor axis of the ellipse is in the range [0.2, 1], and the ellipse area threshold is in the range (0.1 μm). 2 -0.8μm 2 If the ratio of the area occupied by the fluorescence signal of the microtubules in a single platelet to the area of ​​the ellipse is within a preset range of 3, then the microtubules involved in the microtubule image of a single platelet are determined to be "aggregated distribution", and the preset range of 3 is [0.6, 1].

[0032] (3) If the microtubule image of a single platelet does not belong to "regular distribution", "diffuse distribution" and "aggregated distribution", then the microtubules involved in the microtubule image of a single platelet are judged to be "irregular distribution";

[0033] Condition A: Microtubes exceeding a preset ellipse ratio threshold are distributed inside the ellipse; the preset ellipse ratio threshold is above 95%.

[0034] Preferably, the platelet microtubule classification system based on super-resolution images classifies microtubules within a single platelet microtubule image that is "regularly distributed" into subclasses based on the ratio of the outer edge length of the microtubule image within the ellipse to the circumference of the ellipse within a preset range, as follows:

[0035] If the ratio of the length of the outer edge of the microtube image within the ellipse to the circumference of the ellipse is in the range of (0, 0.2), then it is a "cross coil".

[0036] If the ratio of the outer edge length of the microtubule image within the ellipse to the circumference of the ellipse is in (0.75, 1], it is "annular distribution".

[0037] Preferably, the platelet microtubule classification system based on the super-resolution image, the ratio is obtained according to the following formula:

[0038] E = L1 / L2

[0039] In the formula, L1 is the outer edge length of the microtubule image within the ellipse, and L2 is the circumference of the ellipse.

[0040] Preferably, the platelet microtubule classification system based on the super-resolution image, the regular distribution and / or the diffuse distribution, the ratio of the major axis to the minor axis of the ellipse is in [0.5, 1], and the area threshold of the ellipse is in [1.7 μm 2 -12.6 μm 2 ]。

[0041] Preferably, the platelet microtubule classification system based on the super-resolution image, the aggregation distribution, the ratio of the major axis to the minor axis of the ellipse is in [0.5, 1], and the area threshold of the ellipse is in [0.2 μm 2 , 0.8 μm 2 ]。

[0042] Overall, compared with the prior art, the above technical solutions conceived by the present application can achieve the following beneficial effects by classifying single platelet microtubule structures:

[0043] The platelet microtubule classification method based on the super-resolution image provided by the present application classifies single platelet microtubule structures into four categories of "regular distribution", "diffuse distribution", "aggregation distribution" and "irregular distribution" based on the obtained platelet microtubule super-resolution image. According to the classification method, single platelet microtubule structures can be classified, which helps to further understand the role of microtubules in the physiological functions of platelets and the mechanism of microtubule regulation in the physiological functions of platelets, and has important significance for the research and detection of platelet physiological functions, and has certain scientific research value.

[0044] The platelet microtubule classification system based on the super-resolution image provided by the present application can automatically extract and label the complete image of single platelet microtubule structure according to the classification method based on the obtained platelet microtubule super-resolution image, and automatically classify, which can realize the automation of microtubule classification and is beneficial to further research of platelet physiological function. BRIEF DESCRIPTION OF DRAWINGS

[0045] Figure 1 is a process schematic diagram of a platelet microtubule super-resolution image classification system;

[0046] Figure 2 is a platelet microtubule super-resolution fluorescence macrograph;

[0047] Figure 3 is a single platelet microtubule structure classification schematic diagram. DETAILED DESCRIPTION

[0048] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not used to limit the present application. In addition, the technical features involved in the various embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.

[0049] Microtubules, as a unique and important skeleton structure in platelets, play an important role in maintaining the disc shape of resting platelets and assisting the deformation of activated platelets, and are one of the structural bases of platelet maturation and promotion of hemostasis and other physiological processes, and are also a sensitive indicator of their physiological state. In addition, microtubules also undergo different changes in pathological states, such as existing studies observing by electron microscopy that platelet microtubules in patients with giant platelet syndrome are loosely assembled to form a severely disordered marginal band, and platelet microtubules in patients with Wiskott-Aldrich syndrome are thickened or in a disordered and twisted state. It can be seen that changes in microtubule structure can directly affect the function and physiological role of platelets.

[0050] However, current research mainly focuses on the influence of different post-translational modifications on the function of microtubules, and does not classify the structure types of microtubules. Therefore, the classification of microtubule morphology can help to understand how microtubules exert their physiological functions, and assist in the diagnosis and treatment of diseases. With the development of super-resolution imaging technology, the current super-resolution fluorescence microscope breaks through the optical diffraction limit and can meet the accurate identification of microtubule subclasses and fast large field imaging, which can be an effective classification detection means.

[0051] Based on the platelet microtubule super-resolution imaging technology, it is observed from a large number of platelet microtubule super-resolution fluorescence images that microtubules generally exhibit different structures, such as some microtubule structures forming one or more closed ring structures, some of which have no microtubule bands inside the ring, some of which have multiple microtubule bands inside the ring, and some of which have diffuse microtubules inside the ring; or some microtubules are all diffuse and round or approximately round; or some microtubules gather together to form one or more groups; or some microtubules exhibit irregular shapes such as half rings or saddles; it can be seen that platelet microtubules often exhibit different structures, and changes in microtubule structure may be related to information such as changes in platelet physiological function, but there is a lack of related research on these different structures of microtubules in existing research.

[0052] Firstly, the platelet microtubule super-resolution image is processed: the image segmentation algorithm is used to crop the platelet microtubule super-resolution image into a single platelet microtubule immunofluorescence labeled super-resolution image, and binarization is performed to obtain a single platelet microtubule image composed of fluorescent signal pixels and non-fluorescent signal pixels;

[0053] For the single platelet microtubule image, the image segmentation algorithm is used to segment the fluorescent signal pixel area into independent microtubule pixel area, and a single platelet microtubule distribution image is obtained;

[0054] Further, the present study found that classifying the microtubules in a single platelet according to different structures can distinguish different samples, which are classified into four categories: "regular distribution", "diffuse distribution", "aggregated distribution" and "irregular distribution", wherein:

[0055] The regular distribution refers to the state of complete assembly of microtubules in a single platelet, some of which present as crossed coils, and some of which present as rings, and the ring includes the case of no microtubule band inside and one or more microtubule bands inside;

[0056] The diffuse distribution refers to the state of incomplete assembly of microtubules in a single platelet, and the outer contour of the microtubule distribution presents as a circle or an ellipse, and the circle or the ellipse has more than 60% of the inside distributed with diffuse microtubules;

[0057] The aggregated distribution refers to the aggregation of microtubules in a single platelet into a cluster, and the area of a single cluster is (0.1 μm 2 , 0.8 μm 2 ];

[0058] The "irregular" refers to the structure other than the above "regular distribution", "diffuse distribution" and "aggregated distribution".

[0059] Further, for the "regular distribution", the microtubule image within a preset ellipse can be further classified according to the ratio of the outer edge length of the microtubule image within the preset ellipse to the circumference of the ellipse within a preset range, specifically as follows:

[0060] If the ratio of the outer edge length of the microtubule image within the preset ellipse to the circumference of the ellipse is (0, 0.2), it is "crossed coil";

[0061] If the ratio of the outer edge length of the microtubule image within the preset ellipse to the circumference of the ellipse is (0.75, 1], it is "ring distribution".

[0062] Classifying the microtubules into the above categories can well distinguish different samples, and classifying the platelet microtubules is beneficial to further study the physiological function changes of platelets.

[0063] Based on this finding, the present application provides a platelet microtubule classification method based on super-resolution images, which comprises the following steps:

[0064] Obtaining a super-resolution image of microtubule immunofluorescence labeling of a single platelet and binarizing to obtain a microtubule image in a single platelet composed of fluorescent signal pixels and non-fluorescent signal pixels;

[0065] For the microtubule image in the single platelet, the fluorescent signal pixel area is segmented into independent microtubule pixel areas by image segmentation algorithm to obtain a single platelet microtubule distribution image;

[0066] Classifying according to the distribution form of the microtubule in the single platelet, specifically as follows:

[0067] (1) Searching for an ellipse in the microtubule image in the single platelet that satisfies condition A and has the smallest area, the ratio of the major axis to the minor axis of the ellipse is in [0.2, 1], and the ellipse area threshold is in [0.8 μm 2 -30 μm 2 ],

[0068] If the ratio of the area occupied by the microtubule fluorescent signal in the single platelet to the area of the ellipse is in a preset range 1, it is judged that the microtubule involved in the microtubule image in the single platelet is "regularly distributed", and the preset range 1 is in (0, 0.6);

[0069] If the ratio of the area occupied by the microtubule fluorescent signal in the single platelet to the area of the ellipse is in a preset range 2, it is judged that the microtubule involved in the microtubule image in the single platelet is "diffusely distributed", and the preset range 2 is in [0.6, 1];

[0070] Preferably, the regularly distributed and / or diffusely distributed ellipse has a ratio of the major axis to the minor axis in [0.5, 1] and an ellipse area threshold in [1.7 μm 2 -12.6 μm 2 ].

[0071] (2) Searching for an ellipse in the microtubule image in the single platelet that satisfies condition A and has the smallest area, the ratio of the major axis to the minor axis of the ellipse is in [0.2, 1], the ellipse area threshold is in (0.1-0.8 μm 2 ], and the ratio of the area occupied by the microtubule fluorescent signal in the single platelet to the area of the ellipse is in a preset range 3, it is judged that the microtubule involved in the microtubule image in the single platelet is "aggregated distribution", and the preset range 3 is in [0.6, 1];

[0072] Preferably, the aggregated distribution ellipse has a ratio of the major axis to the minor axis in [0.5, 1] and an ellipse area threshold in [0.2 μm 2 , 0.8 μm 2 ].

[0073] (3) the single platelet microtubule image does not belong to "regular distribution", "diffuse distribution" and "aggregated distribution", then the microtubule involved in the single platelet microtubule image is judged as "irregular distribution";

[0074] Condition A: the microtubules exceeding the preset ellipse ratio threshold are distributed inside the ellipse; the preset ellipse ratio threshold is above 95%.

[0075] For the microtubules involved in the single platelet microtubule image of "regular distribution", the microtubule image is subclassified according to the ratio of the outer edge length of the microtubule image inside the ellipse to the circumference of the ellipse in a preset range, specifically as follows:

[0076] If the ratio of the outer edge length of the microtubule image inside the ellipse to the circumference of the ellipse is (0, 0.2), it is "cross coil";

[0077] If the ratio of the outer edge length of the microtubule image inside the ellipse to the circumference of the ellipse is (0.75, 1], it is "ring distribution"; the ratio is calculated according to the following formula:

[0078] E = L1 / L2

[0079] In the formula, L1 is the outer edge length of the microtubule image inside the ellipse, and L2 is the circumference of the ellipse;

[0080] The "ring distribution" refers to the ring distribution of the microtubule image, and the ring includes the case of no microtubule band inside and one or more microtubule bands inside.

[0081] In addition, the present application also provides a platelet microtubule classification system based on super-resolution images, which comprises a platelet microtubule super-resolution image acquisition module, a single platelet microtubule image extraction and labeling module and a classification module;

[0082] The platelet microtubule super-resolution image acquisition module is used to obtain a single platelet microtubule image composed of fluorescent signal pixels and non-fluorescent signal pixels by binarization; and submit to the single platelet microtubule distribution image extraction and labeling module; the resolution of the super-resolution image is below 100 nm, which can detect the platelet microtubule precisely under such high resolution, and can directly detect the change of the platelet microtubule structure, and any one of SIM, STED and STORM super-resolution microscopic imaging technology can be used to obtain the image;

[0083] The platelet microtubule structure labeled by fluorescence is specific, and does not need to be identified by professionals, which is beneficial to the system identification and extraction and labeling of the single platelet microtubule distribution image;

[0084] The single platelet microtubule image extraction and labeling module is configured to, for the single platelet microtubule image, segment the fluorescent signal pixel region into an independent microtubule pixel region by using an image segmentation algorithm, obtain a single platelet microtubule distribution image, and submit the single platelet microtubule distribution image to the classification module.

[0085] The classification module is configured to classify the platelets according to the distribution form of the microtubules, and specifically as follows.

[0086] The platelets are classified according to the distribution form of the microtubules in the single platelet, and specifically as follows.

[0087] (1) search for an ellipse in the single platelet microtubule image that satisfies condition A and has the smallest area, the ratio of the major axis to the minor axis of the ellipse is in [0.2, 1], and the ellipse area threshold is in [0.8 μm 2 -30 μm 2 ],

[0088] If the ratio of the area occupied by the single platelet microtubule fluorescent signal to the area of the ellipse is in a preset range 1, it is determined that the microtubule involved in the single platelet microtubule image is in a “regular distribution”, and the preset range 1 is in (0, 0.6);

[0089] If the ratio of the area occupied by the single platelet microtubule fluorescent signal to the area of the ellipse is in a preset range 2, it is determined that the microtubule involved in the single platelet microtubule image is in a “diffuse distribution”, and the preset range 2 is in [0.6, 1];

[0090] Preferably, the ratio of the major axis to the minor axis of the ellipse in the regular distribution and / or the diffuse distribution is in [0.5, 1], and the ellipse area threshold is in [1.7 μm 2 -12.6 μm 2 ].

[0091] (2) search for an ellipse in the single platelet microtubule image that satisfies condition A and has the smallest area, the ratio of the major axis to the minor axis of the ellipse is in [0.2, 1], the ellipse area threshold is in (0.1 μm 2 -0.8 μm 2 ], and the ratio of the area occupied by the single platelet microtubule fluorescent signal to the area of the ellipse is in a preset range 3, it is determined that the microtubule involved in the single platelet microtubule image is in an “aggregated distribution”, and the preset range 3 is in [0.6, 1];

[0092] Preferably, the ratio of the major axis to the minor axis of the ellipse in the aggregated distribution is in [0.5, 1], and the ellipse area threshold is in [0.2 μm 2 -0.8 μm 2 ].

[0093] (3) the single platelet microtubule image does not belong to "regular distribution", "diffuse distribution" and "aggregated distribution", then the microtubule involved in the single platelet microtubule image is judged as "irregular distribution";

[0094] Condition A: the microtubules exceeding the preset ellipse ratio threshold are distributed inside the ellipse; the preset ellipse ratio threshold is above 95%.

[0095] For the microtubules involved in the single platelet microtubule image of "regular distribution", the microtubule image is subclassified according to the ratio of the outer edge length of the microtubule image inside the ellipse to the circumference of the ellipse in the preset range, specifically as follows:

[0096] If the ratio of the outer edge length of the microtubule image inside the ellipse to the circumference of the ellipse is in (0, 0.2), it is "cross coil";

[0097] If the ratio of the outer edge length of the microtubule image inside the ellipse to the circumference of the ellipse is in (0.75, 1], it is "ring distribution"; the ratio is calculated according to the following formula:

[0098] E = L1 / L2

[0099] In the formula, L1 is the outer edge length of the microtubule image inside the ellipse, and L2 is the circumference of the ellipse.

[0100] The "ring distribution" refers to the ring distribution of the microtubule image, and the ring includes the case of no microtubule band inside and the case of one or more microtubule bands inside.

[0101] The following is an example:

[0102] Example 1: Platelet microtubule classification method based on super-resolution image

[0103] The platelet microtubule classification method based on super-resolution image includes the following steps:

[0104] (1) Obtain platelet microtubule super-resolution image: use anti-αtubulin protein antibody to immunofluorescence label the microtubule in the platelet, and use GI-SIM super-resolution microscope to shoot 40-50 regions to obtain 500 single dispersed multiple platelet super-resolution images, wherein one platelet microtubule super-resolution fluorescence large image is shown as Figure 2 The platelet microtubule super-resolution fluorescence large image is one super-resolution image reconstructed from nine original images according to the principle of SIM super-resolution system;

[0105] (2) Classification: Obtain the microtubule image of a single platelet by binarizing the super-resolution image of the microtubule immunofluorescence labeling of the single platelet; and segment the fluorescent signal pixel region in the microtubule image of the single platelet into independent microtubule pixel regions by using an image segmentation algorithm to obtain the microtubule distribution image of the single platelet; and classify the microtubules according to the distribution form, specifically as follows:

[0106] (1) Search for the ellipse in the microtubule image of the single platelet that meets condition A and has the smallest area, wherein the ratio of the major axis to the minor axis of the ellipse is in the range of [0.2, 1], and the area threshold of the ellipse is in the range of 15.8 μm 2 ;

[0107] If the ratio of the area occupied by the microtubule fluorescent signal in the single platelet to the area of the ellipse is in the range of (0, 0.6), the microtubules involved in the microtubule image of the single platelet are determined to be "regularly distributed".

[0108] If the ratio of the area occupied by the microtubule fluorescent signal in the single platelet to the area of the ellipse is in the range of [0.6, 1], the microtubules involved in the microtubule image of the single platelet are determined to be "diffusely distributed".

[0109] For the microtubules involved in the microtubule image of the single platelet that are "regularly distributed", the microtubules are classified into subclasses according to the ratio of the length of the outer edge of the microtubule image in the ellipse to the circumference of the ellipse, specifically as follows:

[0110] If the ratio of the length of the outer edge of the microtubule image in the ellipse to the circumference of the ellipse is in the range of (0, 0.2), it is "crossed coil".

[0111] If the ratio of the length of the outer edge of the microtubule image in the ellipse to the circumference of the ellipse is in the range of (0.75, 1], it is "ring distribution".

[0112] The ratio is calculated according to the following formula:

[0113] E = L1 / L2

[0114] In the formula, L1 is the length of the outer edge of the microtubule image in the ellipse, and L2 is the circumference of the ellipse. In this embodiment, pixels are used as the unit, the edge operator Canny algorithm is used to detect the edge, and then the number of pixel points on the edge of the microtubule fluorescent image is determined as the length of the outer edge of the microtubule image.

[0115] (2) Search for the ellipse in the microtubule image of the single platelet that meets condition A and has the smallest area, wherein the ratio of the major axis to the minor axis of the ellipse is in the range of [0.2, 1], and the area threshold of the ellipse is in the range of 0.66 μm 2If the ratio of the area of the single platelet microtubule fluorescence signal to the area of the ellipse is in [0.6, 1], it is determined that the microtubule involved in the single platelet microtubule image is "aggregated distribution";

[0116] (3) If the single platelet microtubule image does not belong to "regular distribution", "dispersion distribution" and "aggregated distribution", it is determined that the microtubule involved in the single platelet microtubule image is "irregular distribution";

[0117] Condition A: The microtubules exceeding the preset ellipse ratio threshold are distributed inside the ellipse; the preset ellipse ratio threshold is above 95%.

[0118] Example 2: Platelet microtubule classification system based on super-resolution image

[0119] The platelet microtubule classification system based on super-resolution image comprises a platelet microtubule super-resolution image acquisition module, a single platelet microtubule image extraction and labeling module, and a classification module. The process established by the classification system is as shown in Figure 1 , and the specific process is as follows:

[0120] Obtain the super-resolution image of the platelet microtubule: use anti-αtubulin protein antibody to immunofluorescence label the microtubule in the platelet, and use GI-SIM super-resolution microscope to shoot 40-50 regions to obtain 500 single-dispersed multiple platelet super-resolution images, wherein one platelet microtubule super-resolution fluorescence large image is as shown in Figure 2 .

[0121] The single platelet microtubule image extraction and labeling module: crop the obtained platelet microtubule super-resolution image to obtain a large number of single platelet microtubule fluorescence images, and label the single platelet microtubule fluorescence images of a large number of different samples as "regular distribution", "dispersion distribution", "aggregated distribution" or "irregular distribution" according to the method provided in Example 1, and the distribution type is as shown in Figure 3 .

[0122] For the platelet microtubule of regular distribution, the ratio of the outer edge length of the microtubule image in the obtained preset ellipse to the circumference of the ellipse is in a preset range, and the specific process is as follows:

[0123] If the ratio of the outer edge length of the microtubule image in the ellipse to the circumference of the ellipse is in (0, 0.2), it is "cross coil";

[0124] If the ratio of the outer edge length of the microtubule image in the ellipse to the circumference of the ellipse is in (0.75, 1], it is "ring distribution".

[0125] A large number of different types of microtubule fluorescence pictures manually labeled are trained, verified and tested by using Resnet50, SVM and other algorithms to construct a classification system capable of autonomous cropping and extraction of labeling and labeling classification.

[0126] The present application classifies platelet microtubules, which may help to understand the role of microtubules in the physiological functions of platelets and the mechanism of microtubules in regulating the physiological functions of platelets. In addition, by analyzing the changes of each category of microtubules in the platelet population with decreased or increased adhesion, aggregation and other abilities, the role of microtubules in the adhesion or aggregation of platelets can be understood, and the correlation between microtubule morphology and adhesion or aggregation function can be established, which helps to understand how platelet microtubules change to exert corresponding functions and promote the study of related mechanisms.

[0127] Those skilled in the art will readily understand that the above description is only the preferred embodiment of the present application, and is not intended to limit the present application, and any modifications, equivalent replacements and improvements made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method of classifying platelet microtubules based on super-resolution images, characterized by, The method comprises the following steps: obtaining a super-resolution image of microtubules of a single platelet by immunofluorescence labeling and binarizing to obtain a microtubule image in a single platelet composed of fluorescent signal pixels and non-fluorescent signal pixels; for the microtubule image in the single platelet, the fluorescent signal pixel region is segmented into independent microtubule pixel regions by an image segmentation algorithm to obtain a single platelet microtubule distribution image; the microtubules involved in the single platelet microtubule distribution image are classified according to the distribution form of the microtubules, specifically as follows: (1) search for the ellipse in the microtubule image of the single platelet that satisfies condition A and has the minimum area, the ratio of the major axis to the minor axis is in [0.2, 1], the ellipse area threshold is in [0.8 μm 2 - 30 μm 2 ], (2) search for the ellipse in the microtubule image of the single platelet that satisfies condition A and has the minimum area, the ratio of the major axis to the minor axis is in [0.2, 1], the ellipse area threshold is in [0.8 μm 2 - 30 μm 2 ], (3) search for the ellipse in the microtubule image if the ratio of the area occupied by the microtubule fluorescent signal in the single platelet to the area of the ellipse is in a preset range 1, it is determined that the microtubules involved in the single platelet microtubule image are "regularly distributed", and the preset range 1 is in (0, 0.6); if the ratio of the area occupied by the microtubule fluorescent signal in the single platelet to the area of the ellipse is in a preset range 2, it is determined that the microtubules involved in the single platelet microtubule image are "diffusely distributed", and the preset range 2 is in [0.6, 1]; (2) search for an ellipse in the single platelet microtubule image that satisfies condition A and has the smallest area, the ratio of the major axis to the minor axis of the ellipse is in [0.2, 1], the ellipse area threshold is in (0.1 μm 2 -0.8 μm 2 ], and the ratio of the area occupied by the microtubule fluorescence signal in the single platelet to the area of the ellipse is in a preset range 3, the preset range 3 is in [0.6, 1]; if not, the microtubule involved in the single platelet microtubule image is determined as "random distribution"; and (3) if the ratio of the area occupied by the microtubule fluorescence signal in the single platelet to the area of the ellipse is in a preset range 4, the preset range 4 is in [0.2, 0.6], the microtubule involved in the single platelet microtubule image is determined as "focal distribution". (3) if the single platelet microtubule image does not belong to "regular distribution", "diffuse distribution" and "aggregated distribution", it is determined that the microtubules involved in the single platelet microtubule image are "irregularly distributed"; Condition A: Microtubules exceeding the preset ellipse ratio threshold are distributed inside the ellipse; the preset ellipse ratio threshold is above 95%.

2. The super-resolution image-based platelet microtubule classification method of claim 1, wherein, For the microtubules involved in the single platelet microtubule image of "regular distribution", the ratio of the outer edge fluorescent length of the microtubule image in the ellipse to the circumference of the ellipse is in a preset range for sub-classification, specifically as follows: if the ratio of the outer edge length of the microtubule image in the ellipse to the circumference of the ellipse is in (0, 0.2), it is "cross coil"; if the ratio of the outer edge length of the microtubule image in the ellipse to the circumference of the ellipse is in (0.75, 1], it is "ring distribution".

3. The super-resolution image-based platelet microtubule classification method of claim 2, wherein, The ratio of the outer edge length of the microtubule image in the ellipse to the circumference of the ellipse is calculated according to the following formula: E = L1 / L2 In the formula, L1 is the outer edge length of the microtubule image in the ellipse, and L2 is the circumference of the ellipse.

4. The super-resolution image-based platelet microtubule classification method of claim 1, wherein, The regular and / or random distribution, the ratio of the major axis to the minor axis of the ellipse is in [0.5, 1], the area threshold of the ellipse is in [1.7 pm 2 - 12.6 pm 2 ].

5. The super-resolution image-based platelet microtubule classification method of claim 1, wherein, The aggregation distribution has an ellipse long axis to short axis ratio in [0.5, 1] and an ellipse area threshold in [0.2 pm 2 , 0.8 pm 2 ].

6. A system for classifying platelet microtubules based on super-resolution images, comprising: The method comprises a platelet microtubule super-resolution image acquisition module, a single platelet microtubule image extraction and labeling module, and a classification module. The platelet microtubule super-resolution image acquisition module is used to obtain a single platelet microtubule image composed of fluorescent signal pixels and non-fluorescent signal pixels by binarizing, and submit it to the single platelet microtubule distribution image extraction and labeling module; The single platelet microtubule image extraction and labeling module is used to segment the fluorescent signal pixel region into independent microtubule pixel regions by an image segmentation algorithm for the single platelet microtubule image to obtain a single platelet microtubule distribution image, and submit it to the classification module; The classification module is used to classify the platelets according to the distribution form of the microtubules, specifically as follows: According to the distribution form of the microtubules involved in the single platelet microtubule distribution image, the classification is specifically as follows: (1) search for the ellipse in the microtubule image of the single platelet that satisfies condition A and has the minimum area, the ratio of the major axis to the minor axis is in [0.2, 1], and the area threshold of the ellipse is in [0.8 μm 2 - 30 μm 2 ]; (2) search for the ellipse in the microtubule image of the single platelet that satisfies condition A and has the minimum area, the ratio of the major axis to the minor axis is in [0.2, 1], and the area threshold of the ellipse is in [0.8 μm 2 - 30 μm 2 ]; (3) search for the ellipse in If the ratio of the area of the microtubule fluorescence signal in the single platelet to the area of the ellipse is in a preset range 1, it is determined that the microtubules involved in the single platelet microtubule image are "regularly distributed", and the preset range 1 is in (0, 0.6); If the ratio of the area of the microtubule fluorescence signal in the single platelet to the area of the ellipse is in a preset range 2, it is determined that the microtubules involved in the single platelet microtubule image are "diffusely distributed", and the preset range 2 is in [0.6, 1]; (2) searching for an ellipse in the single-platelet microtubule image that satisfies condition A and has the smallest area, the ratio of the major axis to the minor axis of the ellipse is in [0.2, 1], the ellipse area threshold is in (0.1-0.8 μm 2 ], and the ratio of the area occupied by the microtubule fluorescence signal in the single-platelet microtubule to the area of the ellipse is in a preset range 3, the preset range 3 is in [0.6, 1]; if not, the microtubule involved in the single-platelet microtubule image is determined as "random distribution"; (3) If the single platelet microtubule image does not belong to "regular distribution", "diffuse distribution" and "aggregated distribution", it is determined that the microtubules involved in the single platelet microtubule image are "irregularly distributed"; Condition A: The microtubules exceeding the preset ellipse proportion threshold are distributed in the interior of the ellipse; and the preset ellipse proportion threshold is above 95%.

7. The super-resolution image-based platelet microtubule classification system of claim 6, wherein, For the microtubules involved in the single platelet microtubule image of "regular distribution", the microtubule image is sub-classified according to the ratio of the outer edge length of the microtubule image in the ellipse to the circumference of the ellipse in a preset range, and the specific classification is as follows: If the ratio of the outer edge length of the microtubule image in the ellipse to the circumference of the ellipse is in (0, 0.2), it is "cross coil"; If the ratio of the outer edge length of the microtubule image in the ellipse to the circumference of the ellipse is in (0.75, 1], it is "ring distribution".

8. The super-resolution image-based platelet microtubule classification system of claim 7, wherein, The ratio of the outer edge length of the microtubule image in the ellipse to the circumference of the ellipse is calculated according to the following formula: E = L1 / L2 In the formula, L1 is the outer edge length of the microtubule image in the ellipse, and L2 is the circumference of the ellipse.

9. The super-resolution image-based platelet microtubule classification system of claim 6, wherein, The regular and / or random distribution, the ratio of the major axis to the minor axis of the ellipse is in [0.5, 1], the area threshold of the ellipse is in [1.7 pm 2 - 12.6 pm 2 ].

10. The super-resolution image-based platelet microtubule classification system of claim 6, wherein, The aggregation distribution has an ellipse long axis to short axis ratio in [0.5, 1] and an ellipse area threshold in [0.2 pm 2 , 0.8 pm 2 ].

Images