Transmission Electron Microscopy Identification Method and System for the Isolation Process of Exosomes from Gastric Cancer Tumor Tissue

Abstract

The present invention relates to the field of image recognition, and proposes a transmission electron microscopy identification method and system for the separation process of exosomes in gastric cancer tumor tissues. By using a clustering algorithm to divide the suspected structures in the target lens image, a primary evaluation of the exosome manifestation degree at the edge is carried out based on the edge differentiation degree and the degree of darkness outside the edge of the suspected structure. Then, a secondary evaluation is combined with the background staining change state of the suspected structure, avoiding the influence that the contour of the background impurities is relatively deep in the prior art, resulting in low differentiation between impurities and exosomes, and improving the accuracy of identification. Furthermore, the morphological matching degree of exosomes is calculated according to the internal depression characteristics of the suspected structure, avoiding the problem that the accuracy of the identification result is affected by the difference in the facing situation of different structures in the lens image and the observation perspective, and realizing the auxiliary analysis of the depression characteristics of exosomes under different facing situations of the perspective. The present invention improves the accuracy of transmission electron microscopy identification in the separation process of exosomes in gastric cancer tumor tissues.

Description

Technical Field

[0001] The present invention relates to the field of image recognition, and particularly to a method and system for identifying exosomes in the process of separating gastric cancer tumor tissue by transmission electron microscopy. Background Art

[0002] With the rapid development of tumor immunology research, it is speculated that gastric cancer cells carry lncRNA-ZFAS1 to DC cells through exosomes and regulate their differentiation, thereby inhibiting immune responses and promoting the immune escape of tumor cells. Therefore, in order to determine the specific expression of exosomes in the tumor-related regions of gastric cancer patients, it is crucial to isolate the exosomes from gastric cancer tumor tissue and observe and identify them by transmission electron microscopy.

[0003] In the prior art, in the process of identifying exosomes by transmission electron microscopy, first, negative staining operation needs to be performed on exosomes to improve the distinguishability between exosomes and the background in the lens image. At the same time, according to the fact that the size of exosomes is usually in the range of 30 to 150 nm and their morphological size characteristics often present as concave hemispherical shapes, the size information and morphological information of each structure in the lens image are screened and identified. However, in the actual process of exosome identification, it is often possible that the contours of some impurities in the background are relatively deep, resulting in a low distinguishability between them and exosomes, which affects the accuracy of exosome identification. At the same time, there are differences in the alignment of different structures in the lens image with the observation angle, and the higher the alignment rate of the structure with the viewing angle, the higher the confidence level corresponding to the analysis of its morphological characteristics. Therefore, it further leads to a low accuracy of the identification results obtained by using the consistency of viewing angle confidence in the traditional exosome identification process.

[0004] Therefore, how to design a method for identifying exosomes in the process of separating gastric cancer tumor tissue by transmission electron microscopy to improve the accuracy of the identification results has become an urgent problem to be solved. Summary of the Invention

[0005] Based on this, a method and system for identifying exosomes in the process of separating gastric cancer tumor tissue by transmission electron microscopy proposed by the present invention divides the suspected structures in the target lens image through a clustering algorithm, and conducts a primary evaluation of the exosome manifestation degree on the edge based on the edge distinguishability degree and the degree of darkness outside the edge of the suspected structure, and then combines the background staining change state of the suspected structure for a secondary evaluation, avoiding the influence that the contours of some impurities in the background are relatively deep in the prior art, resulting in a low distinguishability between impurities and exosomes, and improving the accuracy of the identification. Then, the exosome morphology matching degree is calculated according to the internal depression characteristics of the suspected structure, avoiding the problem that the accuracy of the identification results is affected by the difference in the alignment of different structures in the lens image with the observation angle, and realizing the auxiliary analysis of the depression characteristics of exosomes under different viewing angle alignments. The present invention improves the accuracy of identifying exosomes in the process of separating gastric cancer tumor tissue by transmission electron microscopy.

[0006] A transmission electron microscopy identification method for the isolation process of exosomes from gastric cancer tumor tissues proposed by the present invention includes:

[0007] Obtain a target lens image, screen the target lens image according to a clustering algorithm to obtain suspected structures, and perform a primary evaluation process based on the edge characteristics of the suspected structures to obtain the degree of exosome manifestation at the edge. The primary evaluation process is based on the degree of edge differentiation and the degree of darkness on the outer side of the edge of the suspected structure;

[0008] Perform a secondary evaluation process based on the background staining change state of the suspected structure to obtain the degree of exosome tendency. The background staining change state is based on the staining gradient rule in the vicinity of the suspected structure;

[0009] Perform exosome morphology matching calculation based on the internal depression characteristics of the suspected structure to obtain the exosome morphology matching degree;

[0010] Calculate an exosome identification index based on the exosome morphology matching degree to obtain an exosome identification result.

[0011] In summary, according to the above transmission electron microscopy identification method for the isolation process of gastric cancer tumor tissue exosomes, the suspected structures in the target lens image are divided by a clustering algorithm, and the performance degree of exosomes at the edge is evaluated once based on the edge differentiation degree and the degree of darkness outside the edge of the suspected structure. Then, a secondary evaluation is carried out in combination with the background staining change state of the suspected structure, avoiding the influence of the relatively deep contour of impurities in the background part in the prior art, which leads to low differentiation between impurities and exosomes, and improving the accuracy of identification. Furthermore, the morphological matching degree of exosomes is calculated according to the internal depression characteristics of the suspected structure, avoiding the problem that the accuracy of the identification result is affected by the difference in the facing situation of different structures in the lens image and the observation perspective, and realizing the auxiliary analysis of the depression characteristics of exosomes under different facing perspectives. The present invention improves the accuracy of transmission electron microscopy identification in the isolation process of gastric cancer tumor tissue exosomes. Specifically, a target lens image is obtained, the target lens image is screened according to a clustering algorithm to obtain suspected structures, and a primary evaluation process is carried out according to the edge characteristics of the suspected structures to obtain the performance degree of exosomes at the edge. The primary evaluation process is based on the edge differentiation degree and the degree of darkness outside the edge of the suspected structure. A secondary evaluation process is carried out according to the background staining change state of the suspected structure to obtain the tendency degree of exosomes. The background staining change state is based on the staining gradient law in the vicinity of the suspected structure, avoiding the influence of the relatively deep contour of impurities in the background part in the prior art, which leads to low differentiation between impurities and exosomes, and improving the accuracy of identification. The morphological matching calculation of exosomes is carried out according to the internal depression characteristics of the suspected structure to obtain the morphological matching degree of exosomes, avoiding the problem that the accuracy of the identification result is affected by the difference in the facing situation of different structures in the lens image and the observation perspective, and realizing the auxiliary analysis of the depression characteristics of exosomes under different facing perspectives. The exosome identification index is calculated according to the morphological matching degree of exosomes to obtain the exosome identification result. The present invention improves the accuracy of transmission electron microscopy identification in the isolation process of gastric cancer tumor tissue exosomes.

[0012] Further, the step of obtaining the target lens image specifically includes:

[0013] Collect peripheral blood related to the target tumor tissue, isolate serum exosomes using an exosome isolation kit, and resuspend the obtained exosomes in a culture medium to obtain a serum exosome sample. Pipette a first preset volume threshold of the serum exosome sample liquid droplets onto a copper mesh of a transmission electron microscope and let it stand for a first preset time. Then, pipette a first preset volume threshold of a staining solution onto the copper mesh of the transmission electron microscope to perform a staining operation on the serum exosome sample. The staining operation lasts for a second preset time. After staining is completed, turn on the transmission electron microscope to collect an exosome lens image, and perform grayscale conversion processing to obtain a target lens image and upload the target lens image to a data acquisition center, which is used to temporarily store the target lens image before discrimination.

[0014] Further, the step of screening the target lens image according to a clustering algorithm to obtain a suspected structure and performing a primary evaluation process according to the edge feature of the suspected structure to obtain the edge exosome manifestation degree specifically includes:

[0015] Perform Kmeans clustering processing according to the coordinate values and grayscale values of the pixel points in the target lens image. In the Kmeans clustering processing, determine the number of clusters according to the elbow method, and divide the target lens image into multiple clusters;

[0016] Take each cluster as a single suspected structure, and respectively extract the outer edge line of each suspected structure. Calculate the gradient mean of the outer edge line of each suspected structure and the gradient mean of all pixel points in the target lens image according to the Sobel operator;

[0017] Calculate the edge differentiation degree of the suspected structure according to the gradient mean of the outer edge line of the suspected structure and the gradient mean of all pixel points in the target lens image. The edge differentiation degree is positively correlated with the gradient mean of the outer edge line of the suspected structure;

[0018] Calculate the edge exosome manifestation degree of the suspected structure according to the degree of darkness on the outer side of the edge of the suspected structure.

[0019] Further, the step of calculating the edge exosome manifestation degree of the suspected structure according to the degree of darkness on the outer side of the edge of the suspected structure specifically includes:

[0020] Delimit the adjacent range of each suspected structure according to a preset adjacent step length. The adjacent range is the spatial range within a preset adjacent step length from the outside of the suspected structure. If the distance between suspected structures is less than or equal to the preset adjacent step length, the adjacent range is based on the midpoint of the connection line of the centers of the suspected structures;

[0021] Calculate the gray-scale mean within the adjacent range of a single suspected structure and the gray-scale mean within the adjacent ranges of all suspected structures, calculate the degree of darkening outside the edge of the suspected structure based on the gray-scale mean within the adjacent range of the single suspected structure and the gray-scale mean within the adjacent ranges of all suspected structures, and then calculate the degree of exosome manifestation outside the edge of the suspected structure based on the degree of darkening outside the edge of the suspected structure.

[0022] Further, the step of performing a secondary evaluation process according to the background staining change state of the suspected structure to obtain the exosome tendency degree specifically includes:

[0023] Connect the center of each suspected structure to the outermost pixel points in the adjacent range to generate multiple staining gradient paths of the suspected structure. Take the Euclidean distance between the pixel points in each staining gradient path and the center of the suspected structure as the abscissa value, and take the gray-scale value of the pixel points in each staining gradient path as the ordinate value to generate the staining gradient curve corresponding to the suspected structure. Each suspected structure includes multiple staining gradient curves, and each staining gradient path has a uniquely corresponding staining gradient curve;

[0024] Judge whether there is any pixel point on each staining gradient curve whose gray-scale value is lower than that of the adjacent previous pixel point. If there is any pixel point whose gray-scale value is lower than that of the adjacent previous pixel point, then determine that the pixel point is an inverse-order pixel;

[0025] Calculate the number of inverse-order pixels included in each suspected structure, calculate the degree of violation of the staining rule corresponding to the suspected structure based on the number of inverse-order pixels, and then calculate the exosome tendency degree of the suspected structure based on the degree of violation of the staining rule.

[0026] Further, the step of performing exosome morphology matching calculation according to the internal depression feature of the suspected structure to obtain the exosome morphology matching degree specifically includes:

[0027] Extract all the strong edge lines inside the suspected structures according to the Canny edge detection algorithm. When the perspective is directly facing, the strong edge line is a single complete circular-connected domain. When there is a non-direct-facing deviation in the perspective, the strong edge line is a part of a single complete circular-connected domain;

[0028] Screen the suspected structures with a single strong edge line. Start displacement from any endpoint of the strong edge line to pass through all the points on the strong edge line completely. Calculate the Euclidean distance from the geometric center of the suspected structure to each point on the strong edge line. Take the displacement amount as the abscissa value and the Euclidean distance as the ordinate value to generate the internal edge analysis curve of the suspected structure;

[0029] Connect the geometric center of the suspected structure to the two endpoints of the strong edge line and extend it to intersect with the outer edge of the suspected structure, extract the sector-like ring region, calculate the standard deviation of the internal edge analysis curve corresponding to the suspected structure, extract the outer edge line of the sector-like ring region, calculate the external edge analysis curve of the outer edge line of the sector-like ring region, and calculate the standard deviation of the external edge analysis curve;

[0030] Judge the confidence level of the edge smoothness feature according to the standard deviation of the internal edge analysis curve and the standard deviation of the external edge analysis curve. If the confidence level of the edge smoothness feature meets the confidence level requirement, calculate the degree of conformity of the depression feature of the suspected structure;

[0031] Calculate the exosome morphology matching degree of the suspected structure according to the degree of conformity of the depression feature of the suspected structure.

[0032] Further, the step of calculating the exosome identification index according to the exosome morphology matching degree to obtain the exosome identification result specifically includes:

[0033] Calculate the exosome identification index according to the exosome morphology matching degree, and then classify the exosome identification index;

[0034] If the exosome identification index is less than the first preset index threshold, it is determined that the current suspected structure is a non-exosome;

[0035] If the exosome identification index is greater than the second preset index threshold, it is determined that the current suspected structure is an exosome body;

[0036] If the exosome identification index is greater than or equal to the first preset index threshold and less than or equal to the second preset index threshold, the current suspected structure is placed in the list to be observed;

[0037] When the number of suspected structures in the list to be observed is greater than or equal to 50% of the number of suspected structures, adjust the resolution of the transmission lens and collect the lens image again. If the number of suspected structures in the list to be observed is still greater than or equal to 50% of the number of suspected structures after the third identification, manually observe the exosomes in the current identification area.

[0038] A transmission electron microscope identification system for the separation process of exosomes from gastric cancer tumor tissues proposed by the present invention includes:

[0039] A primary evaluation module for obtaining a target lens image, screening the target lens image according to a clustering algorithm to obtain suspected structures, and performing primary evaluation processing according to the edge features of the suspected structures to obtain the degree of exosome manifestation at the edge. The primary evaluation processing is based on the edge differentiation degree and the degree of darkness on the outer side of the edge of the suspected structure;

[0040] A secondary evaluation module, configured to perform secondary evaluation processing based on the background staining change state of the suspected structure to obtain the exosome tendency degree, where the background staining change state is based on the staining gradient rule within the range adjacent to the suspected structure;

[0041] A morphology matching degree calculation module, configured to perform exosome morphology matching calculation based on the internal depression feature of the suspected structure to obtain the exosome morphology matching degree;

[0042] An index discrimination module, configured to calculate an exosome discrimination index based on the exosome morphology matching degree to obtain an exosome discrimination result.

[0043] The present invention also provides a storage medium storing one or more programs, which when executed by a processor implement the transmission electron microscopy discrimination method for the separation process of exosomes from gastric cancer tumor tissues as described above.

[0044] The present invention also provides a computer device, which includes a memory and a processor, where:

[0045] The memory is used to store a computer program;

[0046] The processor is configured to implement the transmission electron microscopy discrimination method for the separation process of exosomes from gastric cancer tumor tissues as described above when executing the computer program stored in the memory. Description of the Drawings

[0047] Figure 1 It is a flowchart of the transmission electron microscopy discrimination method for the separation process of exosomes from gastric cancer tumor tissues proposed in the first embodiment of the present invention;

[0048] Figure 2 It is a structural schematic diagram of the transmission electron microscopy discrimination system for the separation process of exosomes from gastric cancer tumor tissues proposed in the second embodiment of the present invention.

[0049] The following specific embodiments will further illustrate the present invention in conjunction with the above-mentioned drawings. Specific Embodiments

[0050] To facilitate the understanding of the present invention, the present invention will be described more comprehensively below with reference to the relevant drawings. Several embodiments of the present invention are given in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive.

[0051] It should be noted that when an element is referred to as "fixed to" another element, it can be directly on the other element or there may also be an intermediate element. When an element is considered to be "connected to" another element, it can be directly connected to the other element or there may be an intermediate element at the same time. The terms "vertical", "horizontal", "left", "right" and similar expressions used herein are for illustrative purposes only.

[0052] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which this invention belongs. The terms used in the specification of this invention are only for the purpose of describing specific embodiments and are not intended to limit this invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0053] Please refer to Figure 1 , which shows a flowchart of a transmission electron microscopy identification method for the separation process of exosomes from gastric cancer tumor tissues proposed in the first embodiment of the present invention. This transmission electron microscopy identification method for the separation process of exosomes from gastric cancer tumor tissues includes steps S01 to S04, where:

[0054] Step S01: Obtain a target lens image, screen the target lens image according to a clustering algorithm to obtain suspected structures, and perform a primary evaluation process based on the edge characteristics of the suspected structures to obtain the degree of exosome manifestation at the edge;

[0055] It should be noted that in this embodiment, the primary evaluation process is based on the edge differentiation degree and the degree of darkness on the outer side of the edge of the suspected structure. Collect the peripheral blood related to the target tumor tissue, isolate serum exosomes through an exosome isolation kit, and resuspend the obtained exosomes with a culture medium to obtain a serum exosome sample. Drop a first preset volume threshold of the serum exosome sample solution onto a copper grid of a transmission electron microscope and let it stand for a first preset time. Then, take a first preset volume threshold of a staining solution and drop it onto the copper grid of the transmission electron microscope to perform a staining operation on the serum exosome sample. The staining operation lasts for a second preset time. After the staining is completed, turn on the transmission electron microscope to collect an exosome lens image, and perform a grayscale conversion process to obtain a target lens image and upload the target lens image to a data acquisition center, which is used to temporarily store the target lens image before identification;

[0056] In this embodiment, the first preset volume threshold is 10 μL, the first preset time is 2 min, and the second preset time is 1.5 min;

[0057] Since the exosomes have clear contours and are well-distinguished from the external background, clustering is performed on the lens images to obtain each suspected structure. Furthermore, since the essence of negative staining is to stain the background around the exosomes, thus highlighting the morphological structure of the exosomes, this will cause the edges of the exosomes in the lens images to be darkened while the gray level of the background around the exosomes becomes darker. Therefore, the degree of edge distinction of each suspected structure and the degree of darkness outside the edge are analyzed to evaluate the degree of exosome manifestation at the edge;

[0058] According to the coordinate values and gray level values of the pixel points in the target lens image, Kmeans clustering is performed. In the Kmeans clustering, the number of clusters is determined according to the elbow method, and the target lens image is divided into multiple clusters;

[0059] Each of the clusters is regarded as a single suspected structure, and the outer edge lines of each of the suspected structures are extracted respectively. According to the Sobel operator, the gradient mean value of the outer edge line of each suspected structure and the gradient mean value of all pixel points in the target lens image are calculated;

[0060] The degree of edge distinction of the suspected structure is calculated according to the gradient mean value of the outer edge line of the suspected structure and the gradient mean value of all pixel points in the target lens image. The degree of edge distinction is positively correlated with the gradient mean value of the outer edge line of the suspected structure;

[0061] The degree of exosome manifestation at the edge of the suspected structure is calculated according to the degree of darkness outside the edge of the suspected structure;

[0062] If the gradient mean value level of the edge line of the suspected structure is higher, it indicates that the distinction degree between the suspected structure and the background is higher, and thus the suspicion degree of the exosome is higher. In this embodiment, the calculation formula for the degree of edge distinction of the suspected structure is as follows:

[0063] ,

[0064] wherein, represents the degree of edge distinction of the suspected structure, represents the gradient mean value of the outer edge line of the suspected structure, represents the gradient mean value of all pixel points in the target lens image;

[0065] The adjacent range of each suspected structure is delimited according to a preset adjacent step length. The adjacent range is the spatial range within a preset adjacent step length outside the suspected structure. If the distance between suspected structures is less than or equal to the preset adjacent step length, the adjacent range is based on the midpoint of the line connecting the centers of the suspected structures;

[0066] Calculate the grayscale mean within the vicinity of a single suspected structure and the grayscale mean within the vicinity of all suspected structures, calculate the degree of darkening outside the edge of the suspected structure based on the grayscale mean within the vicinity of the single suspected structure and the grayscale mean within the vicinity of all suspected structures, and then calculate the exosome manifestation degree outside the edge of the suspected structure based on the degree of darkening outside the edge of the suspected structure;

[0067] Since negative staining treatment will cause the background grayscale in the vicinity of exosomes to darken, the exosome manifestation degree outside the edge of the suspected structure is calculated in combination with the degree of darkening outside the edge;

[0068] Since the staining degree of each exosome by the staining treatment has little difference, the smaller the difference in the grayscale mean within the vicinity of a certain suspected structure and the grayscale mean within the vicinity of other suspected structures, and the smaller the grayscale mean within the vicinity of this suspected structure, the greater the probability that this suspected structure is an exosome;

[0069] If the edge differentiation degree of the suspected structure is greater and the degree of darkening outside the edge is higher, it indicates that the probability that this suspected structure is an exosome is greater;

[0070] In this embodiment, the preset proximity step is 10 pixel distances, and the calculation formula for the degree of darkening outside the edge of the suspected structure is as follows:

[0071] ,

[0072] Wherein, represents the degree of darkening outside the edge of the suspected structure, represents function, represents the grayscale mean within the vicinity of a single suspected structure, represents the grayscale mean within the vicinity of all suspected structures;

[0073] Define the vicinity of each suspected structure according to the preset proximity step. The vicinity is the spatial range within one preset proximity step outside the suspected structure. If the distance between suspected structures is less than or equal to the preset proximity step, the vicinity is based on the midpoint of the line connecting the centers of the suspected structures;

[0074] Calculate the grayscale mean within the vicinity of a single suspected structure and the grayscale mean within the vicinity of all suspected structures, calculate the degree of darkening outside the edge of the suspected structure based on the grayscale mean within the vicinity of the single suspected structure and the grayscale mean within the vicinity of all suspected structures, and then calculate the exosome manifestation degree outside the edge of the suspected structure based on the degree of darkening outside the edge of the suspected structure.

[0075] The calculation formula for the exosome manifestation degree outside the edge of the suspected structure is as follows:

[0076] ,

[0077] Among them, represents the expression degree of edge exosomes of the suspected structure.

[0078] Step S02: Perform a secondary evaluation process according to the background staining change state of the suspected structure to obtain the exosome tendency degree;

[0079] It should be noted that in this embodiment, the background staining change state is based on the staining gradient law within the vicinity of the suspected structure. When considering staining the exosomes, the farther away from the exosomes, the lighter the staining effect and the lower the degree of background influence. Therefore, the exosome tendency degree of each suspected structure is obtained by combining the background staining situation;

[0080] Connect the center of each suspected structure to the outermost pixel points in the vicinity to generate multiple staining gradient paths of the suspected structure. Take the Euclidean distance between the pixel points and the center of the suspected structure in each staining gradient path as the abscissa value, and take the gray value of the pixel points in each staining gradient path as the ordinate value to generate the staining gradient curve corresponding to the suspected structure. Each suspected structure includes multiple staining gradient curves, and each staining gradient path has a uniquely corresponding staining gradient curve;

[0081] Judge whether there is any pixel point on each staining gradient curve whose gray value is lower than the gray value of the adjacent previous pixel point. If there is any pixel point whose gray value is lower than the gray value of the adjacent previous pixel point, then determine that the pixel point is an inverse-order pixel;

[0082] Calculate the number of inverse-order pixels included in each suspected structure, calculate the degree of violation of the staining rule corresponding to the suspected structure according to the number of inverse-order pixels, and then calculate the exosome tendency degree of the suspected structure according to the degree of violation of the staining rule;

[0083] Negative staining should make the pixel gray value within the vicinity of the suspected structure gradually decrease as the distance from the suspected structure increases. Therefore, the more the number of inverse-order pixels of a certain suspected structure, the less the suspected structure satisfies this change rule, and the lower the possibility that the suspected structure is a stained exosome;

[0084] The calculation formula for the degree of violation of the staining rule of the suspected structure is as follows:

[0085] ,

[0086] Among them, represents the degree of violation of the staining rule of the suspected structure, represents the number of inverse-order pixels;

[0087] If the suspected structure violates the staining rules less and the marginal exosome expression is higher, it means that the suspected structure has a higher exosome tendency;

[0088] The calculation formula for the exosome tropism of suspected structures is as follows:

[0089] ,

[0090] in, Indicates the degree of exosome tropism of the suspected structure.

[0091] Step S03: performing exosome morphology matching calculation based on the internal concave features of the suspected structure to obtain the exosome morphology matching degree;

[0092] It should be noted that in this embodiment, since exosomes have a concave hemispherical feature that tends to be concave toward the center, the deep edge performance of the suspected structure is used to determine whether it has a concave feature inside, and then the suspected structures with concave features are screened out. The degree of concave tendency to the center is combined with the visual angle of the exosome to obtain the degree of concave feature consistency of the exosomes, and the exosome morphological matching degree of each suspected structure is obtained by combining the degree of concave feature consistency and the degree of exosome tendency.

[0093] Because exosomes often appear as a single concave hemisphere, the strong edge line representing the concave edge of the exosome under the transmission electron microscope should be a single line. This strong edge line should be a complete quasi-circular connected domain when the viewing angle is facing the right angle. When the viewing angle is not facing the right angle, the strong edge line should be part of the complete quasi-circular connected domain.

[0094] The strong edge lines inside all suspected structures are extracted using the Canny edge detection algorithm. When the viewing angle is facing the image correctly, the strong edge lines are a single complete quasi-circular connected domain. When the viewing angle is not facing the image correctly, the strong edge lines are part of a single complete quasi-circular connected domain.

[0095] Screen suspected structures with a single strong edge line, start displacement from any endpoint of the strong edge line, and completely pass through all points on the strong edge line. Calculate the Euclidean distance from the geometric center of the suspected structure to each point on the strong edge line, use the displacement as the abscissa value and the Euclidean distance as the ordinate value to generate an internal edge analysis curve for the suspected structure;

[0096] Different from the irregular growth changes of other concave boundaries, the strong edge line of the concave boundary inside the exosome is relatively smooth. Therefore, we connect the geometric center of the suspected structure to the two end points of the strong edge line and extend it to intersect with the outer edge of the suspected structure to extract the fan-shaped ring area.

[0097] Connect the geometric center of the suspected structure to the two endpoints of the strong edge line and extend it to intersect with the outer edge of the suspected structure to extract a fan-shaped ring area, calculate the standard deviation of the internal edge analysis curve corresponding to the suspected structure, extract the outer edge line of the fan-shaped ring area, calculate the outer edge analysis curve of the outer edge line of the fan-shaped ring area, and calculate the standard deviation of the outer edge analysis curve;

[0098] Determining the confidence of the edge smoothing feature according to the standard deviation of the internal edge analysis curve and the standard deviation of the external edge analysis curve, and if the confidence of the edge smoothing feature meets the confidence requirement, calculating the degree of conformity of the concave feature of the suspected structure;

[0099] Calculating the exosome morphology matching degree of the suspected structure according to the degree of conformity of the concave features of the suspected structure;

[0100] Because the strong edge line of the concave boundary inside the exosome presents a relatively smooth feature, and the higher the viewing angle is, the higher the confidence of the concave feature is, and the smaller the standard deviation of the internal edge analysis curve is, the smoother the strong edge line inside the suspected structure is, and the greater the probability that the suspected structure is an exosome;

[0101] The closer the standard deviation corresponding to the outer edge of the suspected structure is to the standard deviation corresponding to the inner edge, the higher the consistency of the change between the inner edge line of the suspected structure and the nearest outer boundary edge, which means that the suspected structure is more likely to be facing the front, and the more concave information appears on the suspected structure at this time, so it can be used as the confidence of the edge smoothing feature. At the same time, the smaller the smoothing feature, the more consistent the concave feature of the suspected structure is with the characteristics of exosomes.

[0102] If the suspected structure has a greater exosome tendency and a higher degree of concave feature conformity, it further indicates that the suspected structure has a higher exosome matching degree;

[0103] The calculation formula for the degree of conformity of the concave features of the suspected structure is as follows:

[0104] ,

[0105] in, Indicates the degree of conformity of the concave features of the suspected structure, Indicates the standard deviation of the internal edge analysis curve corresponding to the suspected structure. In this embodiment, the standard deviation of the internal edge analysis curve corresponding to the suspected structure is used as the edge smoothing feature. represents the standard deviation of the outer edge analysis curve;

[0106] The calculation formula for the exosome morphology matching of suspected structures is as follows:

[0107] ,

[0108] Among them, represents the morphological matching degree of exosomes suspected of being a structure.

[0109] Step S04: Calculate the exosome discrimination index based on the exosome morphological matching degree to obtain the exosome discrimination result;

[0110] It should be noted that in this embodiment, since the size standard of exosomes is 30 to 150 nm, the deviation degree between the size of the suspected structure and the standard size is combined, the exosome discrimination index is calculated according to the exosome morphological matching degree, and then the index classification is performed on the exosome discrimination index;

[0111] If the exosome discrimination index is less than the first preset index threshold, it is determined that the current suspected structure is not an exosome;

[0112] If the exosome discrimination index is greater than the second preset index threshold, it is determined that the current suspected structure is the exosome body;

[0113] If the exosome discrimination index is greater than or equal to the first preset index threshold and less than or equal to the second preset index threshold, the current suspected structure is placed in the list to be observed;

[0114] When the number of suspected structures in the list to be observed is greater than or equal to 50% of the number of suspected structures, the resolution of the transmission lens is adjusted and the lens image is collected again. If the number of suspected structures in the list to be observed is still greater than or equal to 50% of the number of suspected structures after the third identification, the exosomes in the current discrimination area are observed manually;

[0115] The calculation formula of the exosome discrimination index is as follows:

[0116] ,

[0117] Among them, represents the exosome discrimination index, represents the maximum inner diameter length of the suspected structure, represents the reference diameter length. In this embodiment, the reference diameter length is 90;

[0118] In this embodiment, the first preset index threshold is 0.33, and the second preset index threshold is 0.88.

[0119] In summary, according to the above method for identifying exosomes in gastric cancer tumor tissue separation process by transmission electron microscopy, the suspected structures in the target lens image are divided by a clustering algorithm, and the performance degree of exosomes at the edge is evaluated once based on the edge differentiation degree and the degree of darkness outside the edge of the suspected structure. Then, a secondary evaluation is carried out in combination with the background staining change state of the suspected structure, avoiding the influence of the relatively deep contour of the background impurities in the prior art, which leads to low differentiation between impurities and exosomes, and improving the accuracy of identification. Furthermore, the morphological matching degree of exosomes is calculated according to the internal depression characteristics of the suspected structure, avoiding the problem that the accuracy of the identification result is affected by the difference in the facing situation of different structures in the lens image and the observation angle, realizing the auxiliary analysis of the depression characteristics of exosomes under different facing situations of the observation angle. The present invention improves the accuracy of identifying exosomes in gastric cancer tumor tissue separation process by transmission electron microscopy. Specifically, a target lens image is obtained, the target lens image is screened according to a clustering algorithm to obtain suspected structures, and a primary evaluation process is carried out according to the edge characteristics of the suspected structures to obtain the performance degree of exosomes at the edge. The primary evaluation process is based on the edge differentiation degree and the degree of darkness outside the edge of the suspected structure. A secondary evaluation process is carried out according to the background staining change state of the suspected structure to obtain the tendency degree of exosomes. The background staining change state is based on the staining gradient law in the vicinity of the suspected structure, avoiding the influence of the relatively deep contour of the background impurities in the prior art, which leads to low differentiation between impurities and exosomes, and improving the accuracy of identification. The morphological matching calculation of exosomes is carried out according to the internal depression characteristics of the suspected structure to obtain the morphological matching degree of exosomes, avoiding the problem that the accuracy of the identification result is affected by the difference in the facing situation of different structures in the lens image and the observation angle, realizing the auxiliary analysis of the depression characteristics of exosomes under different facing situations of the observation angle. The exosome identification index is calculated according to the morphological matching degree of exosomes to obtain the exosome identification result. The present invention improves the accuracy of identifying exosomes in gastric cancer tumor tissue separation process by transmission electron microscopy.

[0120] Please refer to Figure 2 , which shows a schematic structural diagram of a transmission electron microscopy identification system for exosomes in gastric cancer tumor tissue separation process according to the second embodiment of the present invention. The system includes:

[0121] A primary evaluation module 10, configured to obtain a target lens image, screen the target lens image according to a clustering algorithm to obtain suspected structures, and perform a primary evaluation process according to the edge characteristics of the suspected structures to obtain the performance degree of exosomes at the edge. The primary evaluation process is based on the edge differentiation degree and the degree of darkness outside the edge of the suspected structure;

[0122] The secondary evaluation module 20 is used to perform secondary evaluation processing according to the background staining change state of the suspected structure body to obtain the exosome tendency degree, and the background staining change state is based on the staining gradient rule within the vicinity of the suspected structure body;

[0123] The morphology matching degree calculation module 30 is used to perform exosome morphology matching calculation according to the internal depression characteristics of the suspected structure body to obtain the exosome morphology matching degree;

[0124] The index identification module 40 is used to calculate the exosome identification index according to the exosome morphology matching degree to obtain the exosome identification result.

[0125] The present invention also provides a computer storage medium, on which one or more programs are stored, and when the program is executed by a processor, the above-mentioned transmission electron microscopy identification method for the separation process of exosomes in gastric cancer tumor tissues is implemented.

[0126] The present invention also provides a computer device, including a memory and a processor, wherein the memory is used to store a computer program, and the processor is used to execute the computer program stored on the memory to implement the above-mentioned transmission electron microscopy identification method for the separation process of exosomes in gastric cancer tumor tissues.

[0127] Those skilled in the art can understand that the logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or in combination with these instruction execution systems, apparatus, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.

[0128] More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection part (electronic device) having one or more wirings, a portable computer disk cartridge (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or otherwise processing as appropriate, and then storing it in a computer memory.

[0129] It should be understood that each part of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following technologies well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0130] In the description of this specification, the description referring to terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples", etc. means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0131] The above-described embodiments merely represent several implementation manners of the present invention, and the description thereof is relatively specific and detailed, but should not be construed as a limitation on the scope of the patent of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the appended claims.

Claims

1. A method for identifying exosomes in gastric cancer tumor tissue during the isolation process by transmission electron microscopy, characterized in that, Including: Obtain a target lens image, screen the target lens image according to a clustering algorithm to obtain suspected structures, and perform a primary evaluation process based on the edge features of the suspected structures to obtain the degree of edge exosome manifestation. The primary evaluation process is based on the edge discrimination degree and the degree of darkness tendency outside the edge of the suspected structure; The step of screening the target lens image according to the clustering algorithm to obtain suspected structures, and performing a primary evaluation process based on the edge features of the suspected structures to obtain the degree of edge exosome manifestation specifically includes: Perform Kmeans clustering processing according to the coordinate values and gray values of the pixel points in the target lens image. In the Kmeans clustering processing, determine the number of clusters according to the elbow method, and divide the target lens image into multiple clusters; Take each of the clusters as a single suspected structure, and separately extract the outer edge lines of each of the suspected structures. Calculate the gradient mean of the outer edge line of each suspected structure and the gradient mean of all pixel points in the target lens image according to the Sobel operator; Calculate the edge discrimination degree of the suspected structure according to the gradient mean of the outer edge line of the suspected structure and the gradient mean of all pixel points in the target lens image. The edge discrimination degree is positively correlated with the gradient mean of the outer edge line of the suspected structure; Calculate the degree of edge exosome manifestation of the suspected structure according to the degree of darkness tendency outside the edge of the suspected structure; The step of calculating the degree of edge exosome manifestation of the suspected structure according to the degree of darkness tendency outside the edge of the suspected structure specifically includes: Delimit the adjacent range of each suspected structure according to a preset adjacent step length. The adjacent range is the spatial range within a preset adjacent step length outside the suspected structure. If the distance between suspected structures is less than or equal to the preset adjacent step length, the adjacent range is based on the midpoint of the connection line of the centers of the suspected structures; Calculate the gray mean value within the adjacent range of a single suspected structure and the gray mean value within the adjacent ranges of all suspected structures, so as to calculate the degree of darkness tendency outside the edge of the suspected structure according to the gray mean value within the adjacent range of the single suspected structure and the gray mean value within the adjacent ranges of all suspected structures, and then calculate the degree of edge exosome manifestation of the suspected structure according to the degree of darkness tendency outside the edge of the suspected structure; Perform a secondary evaluation process according to the background staining change state of the suspected structure to obtain the exosome tendency degree. The background staining change state is based on the staining gradient law within the adjacent range of the suspected structure; The step of performing a secondary evaluation process according to the background staining change state of the suspected structure to obtain the exosome tendency degree specifically includes: Connect the center of each suspected structure to the outermost pixels in the adjacent range to generate multiple staining gradient paths of the suspected structures. Take the Euclidean distance between the pixels in each staining gradient path and the center of the suspected structure as the abscissa value, and take the grayscale value of the pixels in each staining gradient path as the ordinate value to generate the staining gradient curve corresponding to the suspected structure. Each suspected structure includes multiple staining gradient curves, and each staining gradient path has a uniquely corresponding staining gradient curve; Determine whether there is any pixel on each staining gradient curve whose grayscale value is lower than that of the adjacent previous pixel. If there is any pixel whose grayscale value is lower than that of the adjacent previous pixel, determine that the pixel is an out-of-order pixel; Calculate the number of out-of-order pixels included in each suspected structure, calculate the degree of violation of the staining rule corresponding to the suspected structure according to the number of out-of-order pixels, and then calculate the exosome tendency degree of the suspected structure according to the degree of violation of the staining rule; Perform exosome morphology matching calculation according to the internal depression characteristics of the suspected structure to obtain the exosome morphology matching degree; The step of performing exosome morphology matching calculation according to the internal depression characteristics of the suspected structure to obtain the exosome morphology matching degree specifically includes: Extract all the strong edge lines inside the suspected structures according to the Canny edge detection algorithm. When the view is directly facing, the strong edge line is a single complete circular-like connected domain. When there is a non-direct facing deviation in the view, the strong edge line is a part of a single complete circular-like connected domain; Select the suspected structures with a single strong edge line. Start displacement from any endpoint of the strong edge line to completely pass through all the points on the strong edge line. Calculate the Euclidean distance from the geometric center of the suspected structure to each point on the strong edge line. Take the displacement amount as the abscissa value and the Euclidean distance as the ordinate value to generate the internal edge analysis curve of the suspected structure; Connect the geometric center of the suspected structure to the two endpoints of the strong edge line and extend them to intersect with the outer edge of the suspected structure. Extract the fan-shaped ring-like area. Calculate the standard deviation of the internal edge analysis curve corresponding to the suspected structure. Extract the outer edge line of the fan-shaped ring-like area, calculate the external edge analysis curve of the outer edge line of the fan-shaped ring-like area, and calculate the standard deviation of the external edge analysis curve; Judge the confidence of the edge smoothness feature according to the standard deviation of the internal edge analysis curve and the standard deviation of the external edge analysis curve. If the confidence of the edge smoothness feature meets the confidence requirement, calculate the degree of conformity of the depression feature of the suspected structure; Calculate the exosome morphology matching degree of the suspected structure according to the degree of conformity of the depression feature of the suspected structure; Calculate the exosome discrimination index according to the exosome morphology matching degree to obtain the exosome discrimination result.

2. The transmission electron microscopy identification method for the exosome isolation process of gastric cancer tumor tissue according to claim 1, characterized in that The step of obtaining the target lens image specifically includes: Collect peripheral blood related to the target tumor tissue, isolate serum exosomes using an exosome isolation kit, and resuspend the obtained exosomes with a culture medium to obtain a serum exosome sample. Drop a first preset volume threshold of the serum exosome sample solution onto a copper mesh of a transmission electron microscope using a pipette and let it stand for a first preset time. Then, drop a first preset volume threshold of a staining solution onto the copper mesh of the transmission electron microscope to perform a staining operation on the serum exosome sample. The staining operation lasts for a second preset time. After staining is completed, turn on the transmission electron microscope to collect an exosome lens image, and perform grayscale conversion processing to obtain a target lens image and upload the target lens image to a data acquisition center, which is used to temporarily store the target lens image before identification.

3. The transmission electron microscopy identification method for the exosome isolation process of gastric cancer tumor tissue according to claim 1, characterized in that, The step of calculating an exosome identification index based on the exosome morphological matching degree to obtain an exosome identification result specifically includes: Calculate an exosome identification index based on the exosome morphological matching degree, and then perform index grading on the exosome identification index; If the exosome identification index is less than a first preset index threshold, determine that the current suspected structure is not an exosome; If the exosome identification index is greater than a second preset index threshold, determine that the current suspected structure is an exosome body; If the exosome identification index is greater than or equal to the first preset index threshold and less than or equal to the second preset index threshold, place the current suspected structure in a list to be observed; When the number of suspected structures in the list to be observed is greater than or equal to 50% of the number of suspected structures, adjust the resolution of the transmission lens and perform another collection of the lens image. If the number of suspected structures in the list to be observed is still greater than or equal to 50% of the number of suspected structures after the third identification, perform manual observation on the exosomes in the current identification area.

4. A transmission electron microscopy identification system for the exosome isolation process of gastric cancer tumor tissues, characterized in that, The system is used to implement the transmission electron microscope identification method for the exosome isolation process of gastric cancer tumor tissue as described in any one of claims 1-3, including: A primary evaluation module, which is used to obtain a target lens image, screen the target lens image according to a clustering algorithm to obtain suspected structures, and perform a primary evaluation process based on the edge features of the suspected structures to obtain the degree of exosome manifestation at the edge. The primary evaluation process is based on the degree of edge differentiation and the degree of darkness on the outer side of the edge of the suspected structure; A secondary evaluation module, which is used to perform a secondary evaluation process according to the background staining change state of the suspected structure to obtain the degree of exosome tendency. The background staining change state is based on the staining gradient rule in the vicinity of the suspected structure; A morphological matching degree calculation module, which is used to perform exosome morphological matching calculation based on the internal depression features of the suspected structure to obtain the exosome morphological matching degree; An index identification module, which is used to calculate an exosome identification index based on the exosome morphological matching degree to obtain an exosome identification result.

5. A storage medium, characterized in that, The storage medium stores one or more programs, which when executed by a processor, implement the transmission electron microscope identification method for the exosome isolation process of gastric cancer tumor tissue as described in any one of claims 1-3.

6. A computer device, characterized in that, The computer device includes a memory and a processor, where: The memory is used to store computer programs; When the processor is used to execute the computer programs stored on the memory, the method for identifying the transmission electron microscope of the exosome separation process of gastric cancer tumor tissues described in any one of claims 1-3 is implemented.

Images