A monitoring method and system for inhibiting the proliferation state of cervical cancer cells by traditional Chinese medicine

Abstract

The present invention relates to the technical field of cell state monitoring, and particularly to a monitoring method and system for inhibiting the proliferation state of cervical cancer cells by traditional Chinese medicine. In order to monitor the variation law of cancer cells during the experiment of inhibiting the proliferation state of cervical cancer cells by traditional Chinese medicine, first, a flat microculture device convenient for 24-hour comprehensive monitoring is set up, so as to facilitate the comprehensive observation of the microscopic images of each cancer cell. Then, by calculating the similarity, comparing it with the similarity threshold, and assigning numbers to the cancer cells, each cancer cell at each moment is corresponded one by one, so that the state change law of each cancer cell during the experiment can be locked, further improving the granularity of the experimental data to better realize the in-depth study of the anti-cervical cancer mechanism of traditional Chinese medicine and the quantitative evaluation of the curative effect. It solves the technical problem that the existing algorithms are difficult to stably track the evolution of the same cell over time, and better realizes the in-depth study of the anti-cervical cancer mechanism of traditional Chinese medicine and the quantitative evaluation of the curative effect.

Description

Technical Field

[0001] The present invention relates to the technical field of cell state monitoring, and particularly to a monitoring method and system for inhibiting the proliferation state of cervical cancer cells by traditional Chinese medicine. Background Art

[0002] In recent years, the anti-cancer potential of traditional Chinese medicine components has attracted much attention. For example, curcumin, as a natural polyphenol compound, has been proven to inhibit the proliferation of cervical cancer cells through mechanisms such as regulating the cell cycle and inducing apoptosis. In the research on cervical cancer, traditional Chinese medicine compounds containing curcumin have shown potential efficacy, but the dynamic evaluation of their effects still faces technical bottlenecks. Existing monitoring methods mostly rely on intermittent microscopic observation or endpoint detection, making it difficult to achieve continuous 24-hour tracking of cell states, resulting in the inability to comprehensively capture the real-time proliferation dynamics, migration behaviors of cancer cells, and response patterns under drug intervention.

[0003] In addition, traditional microscopic image analysis technology has significant deficiencies in dealing with cell overlap, morphological changes, and long-time sequence data association. For example, cells may cause confusion in identity recognition due to division, movement, or occlusion during the culture process. Existing algorithms are difficult to stably track the time evolution of the same cell, thus affecting the accuracy of statistical data. This defect severely restricts the in-depth study of the anti-cervical cancer mechanism of traditional Chinese medicine and the quantitative evaluation of its efficacy. Summary of the Invention

[0004] Aiming at the deficiencies of the existing technology, the present invention provides a monitoring method and system for inhibiting the proliferation state of cervical cancer cells by traditional Chinese medicine, which solves the technical problem that existing algorithms are difficult to stably track the time evolution of the same cell.

[0005] To solve the above technical problems, the present invention provides the following technical solutions:

[0006] A monitoring method for inhibiting the proliferation state of cervical cancer cells by traditional Chinese medicine includes a flat microculture device for culturing cervical cancer cells. The monitoring method specifically includes the following steps:

[0007] S1. Obtain a microscopic image of cancer cells in a direction perpendicular to the flat surface of the microculture device, and preprocess the microscopic image to identify cancer cells;

[0008] S2. Determine whether the current moment is the first moment of the microscopic image;

[0009] If so, obtain the cancer cell data of each cancer cell, and assign different numbers to obtain cancer cell numbers and then end;

[0010] If not, enter step S3;

[0011] S3. Obtain the cancer cell data of each cancer cell at any moment, and calculate the similarity between the cancer cell data at the current moment and the cancer cell rectangular frames of the cancer cell data in history;

[0012] S4. Set a similarity threshold, and determine whether the similarity value is greater than the similarity threshold;

[0013] If so, extract the cancer cell number of the cancer cell data in history with the largest similarity value, and assign it to the cancer cell data at the current moment;

[0014] If not, assign a new cancer cell number to the cancer cell data at the current moment;

[0015] S5. Output monitoring data according to the cancer cell data, cancer cell number, and preset statistical rules.

[0016] Preferably, in step S2, the specific steps for obtaining the cancer cell data are as follows:

[0017] S21. Establish an initial rectangular frame for each cancer cell according to the target detection algorithm, and obtain the cancer cell boundary line;

[0018] S22. Select any side of the rectangular frame as the active side, and mark the remaining sides as the passive sides;

[0019] S23. Make the active side tangent to the cancer cell boundary line to obtain the tangent point, and make the passive side intersect with the cancer cell boundary line and have only one intersection point;

[0020] S24. Move the tangent point along the cancer cell boundary line, and obtain the rectangular frame with the smallest area and only one intersection point between the active side and the cancer cell boundary line, and mark it as the cancer cell rectangular frame;

[0021] S25. Sequentially obtain the two mutually perpendicular sides of each side of the cancer cell rectangular frame, and the first intersection points of the two perpendicular sides and the cancer cell boundary line;

[0022] S26. Obtain the length of the cancer cell boundary line closest to the side between the two first intersection points, and use it as the virtual length of the side;

[0023] S27. According to the side dimensions, angles, positions of the cancer cell rectangular frame, and the virtual lengths of each side, construct an expression of the cancer cell rectangular frame to obtain the cancer cell data; the expression of the cancer cell rectangular frame is as follows:

[0024] ;

[0025] In the above formula, and respectively represent the first mark and the second mark of the cancer cell rectangular frame, and Respectively represent the horizontal and vertical coordinates of the center point of the cancer cell rectangle. and Respectively represent the width and length of the cancer cell rectangle, where Corresponding to the longest side of the cancer cell rectangular box, S is the data set of the virtual length of the cancer cell boundary line corresponding to each box edge, Indicates the minimum angle that the X-axis passes when it is rotated to be parallel to the edge corresponding to any data in the data set. .

[0026] Preferably, in step S24, the following steps are specifically included:

[0027] S241, randomly setting a number of sampling points on the cancer cell boundary line, and moving the cutting point along the boundary line for one circle to obtain a number of first rectangular frames when the cutting point is located at the sampling point;

[0028] S242, selecting a first rectangular frame where the active edge and the cancer cell boundary have two or more intersection points, and marking it as a second rectangular frame;

[0029] S243, rotating the active edge of the second rectangular frame, and determining whether a rectangular frame having only one intersection point between the active edge and the cancer cell boundary line can be obtained;

[0030] If yes, rotate the active edge of the second rectangular frame to obtain a new rectangular frame with the smallest area, and mark it as the first rectangular frame;

[0031] If not, the second rectangular frame and the corresponding sampling points are removed;

[0032] S244, selecting a first rectangular box with the smallest area and having only one intersection point between the active edge and the cancer cell boundary line, and calculating the absolute value of the area difference between the first rectangular box and the first rectangular box obtained last time;

[0033] S245, determining whether the absolute value of the area difference is less than the area difference threshold;

[0034] If yes, the current first rectangular frame is marked as a cancer cell rectangular frame, and then the process goes to step S25;

[0035] If not, proceed to the next step;

[0036] S246, selecting a sampling point corresponding to the first rectangular frame with the smallest area, and marking two sampling points adjacent to the sampling point as segmentation points;

[0037] S247, setting a number of sampling points again between the two segmentation points, and moving the cutting point along the cancer cell boundary line between the two segmentation points, and then returning to step S242.

[0038] Preferably, in step S245, the area difference threshold is 1.0% of the area of the first rectangular box with the smallest area obtained last time.

[0039] Preferably, in step S241, the steps for setting the sampling points are as follows:

[0040] S2411. Randomly set a number of initial sampling points on the cancer cell boundary line;

[0041] S2412. Obtain each convex hull on the cancer cell boundary line;

[0042] S2413. Calculate the ratio of the depth to the arc length of each convex hull in turn;

[0043] S2414. Set a depth-to-length ratio threshold, and remove the convex hulls with a ratio of depth to arc length less than the depth-to-length ratio threshold to obtain standard convex hulls;

[0044] S2415. Set sampling points at the vertices, starting points, and ending points of the standard convex hulls.

[0045] Preferably, the number of the initial sampling points is greater than or equal to 12.

[0046] Preferably, in step S3, it specifically includes the following steps:

[0047] S31. Rotate the angles of the cancer cell rectangular frames of the current moment and the cancer cell data in history around their center points to the same angle, and this angle is a multiple of 90°;

[0048] S32. Calculate the IOU values between the cancer cell data at the current moment after conversion and the cancer cell rectangular frames corresponding to the cancer cell data in history; the calculation formula of the IOU value is:

[0049] ;

[0050] In the above formula, A and B respectively represent the cancer cell rectangular frame of the cancer cell data at the current moment and any cancer cell data in history, and respectively represent the abscissa and ordinate of the center point of the cancer cell rectangular frame A, and respectively represent the width and length of the cancer cell rectangular frame A, and respectively represent the abscissa and ordinate of the center point of the cancer cell rectangular frame B, and respectively represent the width and length of the object B, and respectively represent the horizontal and vertical overlap degrees of the cancer cell rectangular frame A and the cancer cell rectangular frame B, Represents the overlapping area of cancer cell rectangle A and cancer cell rectangle B. and respectively represent the areas of cancer cell rectangle A and cancer cell rectangle B. represents the IOU value of cancer cell rectangle A and cancer cell rectangle B.

[0051] S33. Calculate the similarity between the cancer cell data at the current moment and any cancer cell data in history based on the IOU value and cancer cell data; the formula for similarity is:

[0052] ;

[0053] In the above formula, represents the similarity, and respectively represent the first coefficient and the second coefficient, represents the standard deviation of the difference between the data of the virtual length in the dataset of cancer cell rectangle A and the corresponding data in the dataset of the virtual length in cancer cell rectangle B.

[0054] Preferably, in step S4, it specifically includes the following steps:

[0055] S41. Set a similarity threshold and determine whether the similarity value is greater than the similarity threshold;

[0056] If so, extract the cancer cell number of the cancer cell data in history with the largest similarity value and mark it as the candidate number of the current cancer cell data, and then proceed to the next step;

[0057] If not, proceed to step S43;

[0058] S42. Determine whether the cancer cell number that is the same as the candidate number of the current cancer cell data appeared at the previous moment;

[0059] If so, use this candidate number as the cancer cell number of the current cancer cell data, and then proceed to step S5;

[0060] If not, proceed to the next step;

[0061] S43. Determine whether there are intersections between the cancer cell border of the current cancer cell data and the cancer cell data of other cells;

[0062] If so, mark this cancer cell data as the cancer cell data to be classified and do not assign a cancer cell number to this cancer cell data, and then proceed to step S5;

[0063] If not, proceed to the next step;

[0064] S44. Determine whether the similarity value between the current cancer cell data and the cancer cell data in history is greater than the similarity threshold;

[0065] If so, extract the cancer cell number of the historical cancer cell data with the largest similarity value, mark it as the candidate number of the current cancer cell data, and then proceed to the next step;

[0066] If not, assign a new cancer cell number to the cancer cell data at the current moment, and then proceed to the next step;

[0067] S45. Starting from the current moment, calculate the similarity values between the cancer cell data at each moment and the cancer cell rectangle corresponding to the cancer cell data to be classified at the previous moment in sequence;

[0068] S46. Determine whether the similarity value is greater than the similarity threshold;

[0069] If so, extract the cancer cell number of the cancer cell data with the largest similarity value and assign it to the cancer cell data to be classified corresponding to the previous moment, and then proceed to the next step;

[0070] If not, proceed to the next step.

[0071] Preferably, the cancer cell number is a positive integer, and each time a new cancer cell number is assigned to the cancer cell data at the current moment, the new cancer cell number is 1 greater than the largest cancer cell number value that has been generated.

[0072] The present invention also provides a monitoring system for inhibiting the proliferation state of cervical cancer cells by traditional Chinese medicine, including a processor and a memory. The memory is used to store a computer program, and when the computer program is executed by the processor, the monitoring method for inhibiting the proliferation state of cervical cancer cells by traditional Chinese medicine is implemented.

[0073] By means of the above technical solution, the present invention provides a monitoring method and system for inhibiting the proliferation state of cervical cancer cells by traditional Chinese medicine, which at least has the following beneficial effects:

[0074] 1. In order to monitor the change law of cancer cells during the experiment of inhibiting the proliferation state of cervical cancer cells by traditional Chinese medicine, the present invention first sets up a flat microculture device convenient for 24-hour comprehensive monitoring, so as to facilitate the comprehensive observation of the microscopic images of each cancer cell. Then, by calculating the similarity, comparing it with the similarity threshold, and assigning cancer cell numbers, each cancer cell at each moment is corresponding, so as to be able to lock the state change law of each cancer cell during the experiment, further improving the granularity of the experimental data, so as to better realize the in-depth research on the anti-cervical cancer mechanism of traditional Chinese medicine and the quantitative evaluation of the curative effect.

[0075] 2. In order to prevent the situation where, when the number of cancer cells is huge, the probability of having two or more cancer cells with similar morphologies is relatively high, which may lead to incorrect similarity determination and thus incorrect tracking of cancer cells, the present invention combines a rectangular frame with the virtual length corresponding to the frame edge to comprehensively evaluate the similarity of cancer cells from two aspects, thereby being more accurate.

[0076] 3. In view of the limited computing power of the hardware in the laboratory, while ensuring computing power and data accuracy, the present invention reduces the requirements for the hardware performance of the laboratory and makes it more universal by setting the initial number of sampling points and further subdividing and setting sampling points at key positions.

[0077] 4. The present invention realizes the recognition of cancer cells that reappear after occlusion and newly proliferated cancer cells by further setting steps S41 - S46, and assigns numbers to the cancer cells, thereby realizing the tracking of the morphological and trajectory changes of various cancer cells.

[0078] 5. By making the cancer cell numbers positive integers, and when assigning a new cancer cell number to the cancer cell data at the current moment each time, the new cancer cell number is 1 greater than the largest cancer cell number value that has already been generated, the present invention can intuitively understand the variation laws of the number of newly proliferated cancer cells, the number of dead cancer cells, and the changes in the volume of cancer cells, etc. according to the numerical changes of the cancer cell numbers. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] The drawings described herein are used to provide a further understanding of the present application, form a part of the present application, and the illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation to the present application. In the drawings:

[0080] Figure 1 is a flowchart of the monitoring method for the inhibitory effect of traditional Chinese medicine on the proliferation state of cervical cancer cells in the present invention;

[0081] Figure 2 is a diagram of cervical cancer cells;

[0082] Figure 3 is a schematic diagram of the virtual length. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0083] In order to make the above - mentioned objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below with reference to the drawings and specific embodiments. Thereby, the implementation process of how the present application uses technical means to solve technical problems and achieve technical effects can be fully understood and implemented accordingly.

[0084] Curcumin is a polyphenolic substance extracted from traditional Chinese medicine plants such as Curcuma longa, Curcuma zedoaria, and Curcuma aromatica. It has attracted much attention due to its multiple pharmacological effects such as anti-inflammatory, antioxidant, anti-tumor, and antibacterial. Curcumin has been found to have a significant inhibitory effect on tumor development in the disease progression of cervical cancer. In recent years, a large number of in vitro and in vivo studies have found that curcumin can play a therapeutic effect in the treatment of cervical cancer through various pathways such as promoting apoptosis, inhibiting the proliferation, metastasis, and invasion of tumor cells, inhibiting the integration of HPV with host cells, and inducing autophagy of tumor cells. The mechanism of action of curcumin on cervical cancer is extensive and it may become a new type of anti-tumor drug in the future. However, due to the poor water solubility, fast decomposition and metabolism, and low bioavailability of curcumin, there is still much room for exploration in the research of curcumin derivatives and new preparations. Existing studies have found that curcuminoids have a therapeutic effect on cervical cancer. Reports show that curcumin may bind to the E6 protein. Therefore, it is necessary to explore the role of curcumin in cervical cancer sensitization and its possible molecular mechanism. We put forward the following hypothesis and demonstrated it: Curcumin competitively binds to the ZNF143 protein with LOC644656, resulting in the inability of ZNF143 to be guided by LOC644656 to bind to the E6-AP promoter region, blocking the ubiquitination of p53 protein by the LOC644656 / ZNF143 / E6-AP signaling axis, and sensitizing cervical cancer to cisplatin.

[0085] Therefore, in order to more stably track the time evolution of the same cancer cells and improve the accuracy of statistical data, the present invention provides a monitoring method for inhibiting the proliferation state of cervical cancer cells by traditional Chinese medicine, as Figure 1 shown, to achieve in-depth research on the anti-cervical cancer mechanism of traditional Chinese medicine and quantitative evaluation of the curative effect, including a flat microculture device for culturing cervical cancer cells. The monitoring method specifically includes the following steps:

[0086] S1. Obtain a microscopic image of cancer cells in a direction perpendicular to the flat surface of the microculture device, and preprocess the microscopic image to identify cancer cells. The thickness of the culture space in the microculture device is between the minimum value of the diameter of a single cervical cancer cell and the maximum value of the diameter of two cervical cancer cells, generally set between 20 micrometers and 60 micrometers. The diameter range of a single cervical cancer cell is usually between 20 micrometers and 30 micrometers. The common microculture device is mainly a microfluidic chip, which is processed to form a horizontally extending culture space inside. During the culture process, the cancer cells inside are observed in the vertical direction. The preprocessing methods include the support vector machine method, methods based on the watershed algorithm, GVF Snake model, and rough set, etc.

[0087] S2. Judge whether the current moment is the first moment of the microscopic image;

[0088] If so, obtain the cancer cell data of each cancer cell, assign different numbers respectively to obtain the cancer cell numbers, and then end;

[0089] If not, go to step S3;

[0090] The acquisition of cancer cell data is convenient for tracking cancer cells. At each moment, the cancer cell data at this moment is compared with the cancer cell data at the previous moment, so as to judge the corresponding relationship between the cancer cells at this moment and the cancer cells at the previous moment, and realize the tracking of cancer cells. However, in the case of a large number of cancer cells, the probability of cancer cells with similar morphologies appearing is relatively high, so the situation of incorrect similarity determination may occur, resulting in incorrect tracking of cancer cells. Therefore, the acquisition method of cancer cell data is refined to reduce the probability of misidentification. The specific steps of cancer cell data acquisition are as follows:

[0091] S21. Establish initial rectangular frames for each cancer cell according to the target detection algorithm, and obtain the cancer cell boundary lines. The target detection algorithm can adopt common algorithms such as the YOLO algorithm;

[0092] S22. Select any side of the rectangular frame as the active side, and mark the remaining sides as the passive sides;

[0093] S23. Make the active side tangent to the cancer cell boundary line to obtain the tangent point, and make the passive side intersect with the cancer cell boundary line and have only one intersection point;

[0094] S24. Move the tangent point along the cancer cell boundary line, and obtain the rectangular frame with the smallest area and with only one intersection point between the active side and the cancer cell boundary line, and mark it as the cancer cell rectangular frame. By obtaining the cancer cell rectangular frame with the smallest area, the correlation degree between the size of the cancer cell rectangular frame and the morphological characteristics of the cancer cell itself can be strengthened, which is convenient for subsequent determination according to the similarity. The acquisition of the cancer cell rectangular frame specifically includes the following steps:

[0095] S241. Randomly set a number of sampling points on the cancer cell boundary line, and move the tangent point along the boundary line for one week to obtain a number of first rectangular frames when the tangent point is located at the sampling points. The setting of the sampling points needs to meet both the computing power requirements of the monitoring system hardware and the requirements of the accuracy of the monitoring data. The setting steps of the sampling points are as follows:

[0096] S2411. Randomly set a number of initial sampling points on the cancer cell boundary line. Through a large number of experiments, it is shown that at least 8 directions of detection are required for the common morphological characteristics (depressions, lobulated edges) of cervical cancer cells. When the number of sampling points is equal to 12, more than 85% of the boundary information can be maintained. Therefore, it is best to make the number of initial sampling points greater than or equal to 12.

[0097] S2412. Obtain each convex hull on the boundary line of cancer cells. The points on the convex hull will cause significant changes in the size and position of the rectangular box. Therefore, these points are regarded as key positions to establish a rectangular box in the subsequent steps.

[0098] S2413. Calculate the ratio of the depth to the arc length of each convex hull in sequence to eliminate noise data.

[0099] S2414. Set a depth-to-arc-length ratio threshold, and eliminate the convex hulls whose ratio of depth to arc length is less than the threshold to obtain standard convex hulls. The depth-to-arc-length ratio threshold is generally set to 0.05.

[0100] S2415. Set sampling points at the vertices, starting points, and ending points of the standard convex hulls.

[0101] S242. Screen out the first rectangular boxes where the active side has two or more intersection points with the boundary line of cancer cells, and mark them as the second rectangular boxes.

[0102] S243. Rotate the active side of the second rectangular box and determine whether a rectangular box can be obtained where the active side has only one intersection point with the boundary line of cancer cells.

[0103] If so, rotate the active side of the second rectangular box to obtain a new rectangular box with the smallest area, and mark it as the first rectangular box.

[0104] If not, eliminate the second rectangular box and the corresponding sampling points.

[0105] S244. Select the first rectangular box with the smallest area and where the active side has only one intersection point with the boundary line of cancer cells, and calculate the absolute value of the area difference between it and the previously obtained first rectangular box. The reason for calculating the absolute value is mainly that the volume of cancer cells may increase or decrease during the culture process. Therefore, the influence of the positive or negative sign of the result is removed by calculating the absolute value.

[0106] S245. Determine whether the absolute value of the area difference is less than the area difference threshold; according to experiments, the change in the area of the first rectangular box of cervical cancer cells at two adjacent times does not exceed 1.0% of the area of the first rectangular box at the previous time. Therefore, the area difference threshold is set to 1.0% of the area value of the previously obtained first rectangular box.

[0107] If so, mark the current first rectangular box as the cancer cell rectangular box, and then proceed to step S25.

[0108] If not, proceed to the next step.

[0109] S246. Select the sampling points corresponding to the first rectangular box with the smallest area, and mark the two sampling points adjacent to this sampling point as the segmentation points. At this time, the sampling points corresponding to the final rectangular box with the smallest area are located on the boundary line of cancer cells between the two segmentation points.

[0110] S247. Set several sampling points again between the two segmentation points for further detailed sampling, and move the tangent point along the cancer cell boundary line between the two segmentation points. Then return to step S242. During this process, if the computing power of the hardware is sufficient, a large number of sampling points can be set at one time, so that there is no need to further set sampling points in subsequent steps S241 - S247.

[0111] S25. Sequentially obtain two mutually perpendicular sides of each side of the cancer cell rectangular frame, and the first intersection points of the two perpendicular sides with the cancer cell boundary line.

[0112] S26. Obtain the length of the cancer cell boundary line closest to the side of the cancer cell rectangular frame between the two first intersection points, and use it as the virtual length of the side. As shown in the figure, the two mutually perpendicular sides of the A side are the B side and the D side respectively. The two first intersection points of the B side and the D side with the cancer cell boundary line are the intersection point 3 and the intersection point 1 respectively. There are two segments of the cancer cell boundary line between the intersection point 1 and the intersection point 3. Take the length of the cancer cell boundary line closest to the side (i.e., the A side) as the virtual length of the A side. By this method, the virtual lengths of the A, B, C, and D sides can be obtained. Figure 3 Shown as, the two mutually perpendicular sides of the A side are the B side and the D side respectively. The two first intersection points of the B side and the D side with the cancer cell boundary line are the intersection point 3 and the intersection point 1 respectively. There are two segments of the cancer cell boundary line between the intersection point 1 and the intersection point 3. Take the length of the cancer cell boundary line closest to the side (i.e., the A side) as the virtual length of the A side. By this method, the virtual lengths of the A, B, C, and D sides can be obtained.

[0113] S27. According to the side dimensions, angles, positions of the cancer cell rectangular frame, and the virtual lengths of each side, construct an expression of the cancer cell rectangular frame to obtain cancer cell data. The expression of the cancer cell rectangular frame is as follows:

[0114] ;

[0115] In the above formula, and respectively represent the first label and the second label of the cancer cell rectangular frame. and respectively represent the abscissa and ordinate of the center point of the cancer cell rectangular frame. and respectively represent the width and length of the cancer cell rectangular frame. Among them, corresponds to the longest side in the cancer cell rectangular frame. S is the dataset of the virtual lengths of the cancer cell boundary lines corresponding to each side. One to four virtual lengths corresponding to the sides can be selected as the data in this dataset. represents the minimum angle passed when the X - axis rotates to be parallel to any side corresponding to the data in this dataset. , for example, the X - axis rotates to be parallel to the L side with the maximum length of the cancer cell boundary line.

[0116] S3. Obtain the cancer cell data of each cancer cell at any moment, and calculate the similarity between the cancer cell data at the current moment and the cancer cell bounding boxes of each cancer cell data in history. Since there are a large number of cancer cells in the microculture device, if only the size of the cancer cell bounding box is used to judge the similarity, the probability of misjudgment is relatively high. At this time, in order to improve its accuracy, further combine the dataset of the virtual lengths corresponding to the sides of the cancer cell bounding box for similarity calculation, and incorporate the morphological characteristics of the boundary line of the cancer cell into the calculation factors of similarity. The specific steps are as follows:

[0117] S31. Rotate the angles of the cancer cell bounding boxes of the current moment and each cancer cell data in history around their center points to the same angle, and this angle is a multiple of 90°, so that the sides of the cancer cell bounding box are either parallel or perpendicular to the x-axis, thus facilitating the accurate calculation of the IOU value in the subsequent steps;

[0118] S32. Calculate the IOU value between the cancer cell data at the converted current moment and the cancer cell bounding boxes corresponding to each cancer cell data in history. In the actual monitoring process, in order to reduce the calculation amount, only compare the cancer cell data at the current moment with the cancer cell data in history within a certain range centered on this cancer cell and calculate the IOU value. Through the coordinates of the center of the current cancer cell and the preset radius, select the cancer cells in history whose distance from the current cancer cell is within a certain radius. The calculation formula of the IOU value is:

[0119] ;

[0120] In the above formula, A and B respectively represent the cancer cell bounding box corresponding to the cancer cell data at the current moment and any cancer cell data in history, and respectively represent the abscissa and ordinate of the center point of the cancer cell bounding box A, and respectively represent the width and length of the cancer cell bounding box A, and respectively represent the abscissa and ordinate of the center point of the cancer cell bounding box B, and respectively represent the width and length of the object B, and respectively represent the horizontal and vertical overlap degrees of the cancer cell bounding box A and the cancer cell bounding box B, represents the overlapping area of the cancer cell bounding box A and the cancer cell bounding box B, and respectively represent the areas of the cancer cell bounding box A and the cancer cell bounding box B, represents the IOU value of the cancer cell bounding box A and the cancer cell bounding box B;

[0121] S33. Calculate the similarity between the cancer cell data at the current moment and any cancer cell data in history based on the IOU value and cancer cell data; the formula for calculating the similarity is:

[0122] ;

[0123] In the above formula, represents the similarity, and represent the first coefficient and the second coefficient respectively, represents the standard deviation of the difference between the data in the dataset of the virtual length in cancer cell rectangle A and the corresponding data in the dataset of the virtual length in cancer cell rectangle B, that is, first calculate the difference between the data in the dataset of the virtual length in cancer cell rectangle A and the corresponding data in the dataset of the virtual length in cancer cell rectangle B, and then calculate the standard deviation of this difference, which is .

[0124] S4. Set a similarity threshold and determine whether the similarity value is greater than the similarity threshold;

[0125] If so, extract the cancer cell number of the cancer cell data in history with the largest similarity value and assign it to the cancer cell data at the current moment;

[0126] If not, assign a new cancer cell number to the cancer cell data at the current moment; During the monitoring process, it is found that some cancer cells will disappear due to being blocked by another cancer cell during the monitoring process, and may even move with the blocking cancer cell. After a period of time, the blocked cancer cell may reappear, or the cancer cell may directly die during the blocked period. It is necessary to identify this type of cancer cell to ensure the consistency of the cancer cell numbers of the same cancer cell during the monitoring process. The specific steps are as follows:

[0127] S41. Set a similarity threshold and determine whether the similarity value between the current cancer cell data and the cancer cell data in history is greater than the similarity threshold. The similarity threshold is generally set between 0.5 - 0.9. Specifically, it can be based on the clarity of the microscopic image. When the clarity is large, the cancer cell data will be more diverse. At this time, the similarity threshold can be set around 0.8 - 0.9 to achieve better differentiation. When the clarity is small, the data diversity is less, and the values of the cancer cell data are less accurate. Therefore, it is necessary to make the similarity threshold smaller, generally around 0.5 - 0.7, to avoid the inability to match the cancer cell numbers;

[0128] If so, extract the cancer cell number of the cancer cell data in history with the largest similarity value and mark it as the candidate number for the current cancer cell data, and then proceed to the next step;

[0129] If not, it indicates that the cancer cells corresponding to the cancer cell data may be newly proliferated, and proceed to step S43 for judgment;

[0130] S42. Determine whether a cancer cell number identical to the candidate number of the current cancer cell data appeared at the previous moment;

[0131] If so, use this candidate number as the cancer cell number of the current cancer cell data. At this time, it indicates that the current cancer cell was not blocked at the previous moment, and then proceed to step S5;

[0132] If not, it indicates that the current cancer cell was blocked at the previous moment, and at this time proceed to the next step;

[0133] S43. Determine whether there are intersections between the cancer cell border of the current cancer cell data and the cancer cell data of other cancer cells, that is, whether there are intersections between the boundary lines of the cancer cells being judged currently and the boundary lines of other cancer cells at the current moment. As Figure 2 shown, it is the intersections of the boundary lines between multiple cancer cells;

[0134] If so, mark this cancer cell data as candidate cancer cell data for classification, and do not assign a cancer cell number to this cancer cell data. At this time, since the cancer cells corresponding to this cancer cell data were blocked at the previous moment and reappeared at the current moment, and this cancer cell also came into contact with and was squeezed by other cancer cells, it is not clear whether it is a newly proliferated cancer cell or an existing cancer cell before, and it is necessary to continue to observe and judge in the subsequent process. Then proceed to step S5;

[0135] If not, proceed to the next step;

[0136] S44. Determine whether the similarity value between the current cancer cell data and the cancer cell data in history is greater than the similarity threshold;

[0137] If so, extract the cancer cell number of the cancer cell data in history with the largest similarity value, and mark it as the candidate number of the current cancer cell data, so as to complete the re - tracking of the blocked cancer cells when they appear after being blocked. Then proceed to the next step;

[0138] If not, assign a new cancer cell number to the cancer cell data at the current moment, which indicates that this cancer cell is newly proliferated. Then proceed to the next step;

[0139] S45. Starting from the current moment, calculate the similarity values between the cancer cell data at each moment and the cancer cell rectangle corresponding to the candidate cancer cell data for classification at the previous moment in sequence;

[0140] S46. Determine whether the similarity value is greater than the similarity threshold;

[0141] If so, extract the cancer cell number of the cancer cell data with the maximum similarity value, and assign it to the cancer cell data to be classified corresponding to the previous moment, and then proceed to the next step;

[0142] If not, proceed to the next step.

[0143] In addition, for the convenience of subsequent counting, it is best to make the cancer cell number a positive integer. And each time a new cancer cell number is assigned to the cancer cell data at the current moment, the new cancer cell number is 1 greater than the largest cancer cell number value that has been generated. Thus, it is possible to know how many new cancer cells have proliferated and how many have died based on the change in the cancer cell number value, and to know the change rule of the cancer cell volume based on the cancer cell number of the cancer cells.

[0144] S5. Output monitoring data according to the cancer cell data, cancer cell number, and preset statistical rules, such as counting the number of dead cancer cells, the number of newly proliferated cancer cells, and the change rule of the cancer cell volume, etc.

[0145] The present invention also provides a monitoring system for the inhibitory state of traditional Chinese medicine on the proliferation of cervical cancer cells, including a processor and a memory. The memory is used to store a computer program, and when the computer program is executed by the processor, it realizes the monitoring method for the inhibitory state of traditional Chinese medicine on the proliferation of cervical cancer cells.

[0146] Those of ordinary skill in the art can understand that all or part of the steps in implementing the above method embodiments can be completed by instructing relevant hardware through a program. Therefore, this application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, this application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0147] Each embodiment in this specification is described in a progressive manner. The key point of each embodiment is to illustrate the differences from other embodiments. The same or similar parts among the embodiments can be referred to each other. For the above embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can refer to the partial description of the method embodiments.

[0148] The above embodiments have introduced the present invention in detail. Specific examples are used in this article to elaborate on the principle and implementation of the present invention. The description of the above embodiments is only used to help understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, based on the idea of the present invention, there will be changes in the specific implementation and application scope. In summary, the content of this specification should not be construed as a limitation to the present invention.

Claims

1. A monitoring method for the inhibitory effect of traditional Chinese medicine on the proliferation state of cervical cancer cells, including a flat microculture device for culturing cervical cancer cells, characterized in that, The monitoring method specifically includes the following steps: S1. Obtain a microscopic image of cancer cells in a direction perpendicular to the flat surface of the microculture device, and preprocess the microscopic image to identify cancer cells; S2. Determine whether the current moment is the first moment of the microscopic image, and the first moment is the initial moment of the microscopic image; If so, obtain the cancer cell data of each cancer cell, and assign different numbers to obtain cancer cell numbers, and then return to step S1 to continue obtaining the microscopic image of the next moment; the cancer cell data includes the cancer cell rectangular frame established for each cancer cell according to the target detection algorithm, and the frame edge size, angle, position of the cancer cell rectangular frame, and the virtual length of each frame edge. The virtual length is obtained by acquiring two mutually perpendicular sides of each frame edge of the cancer cell rectangular frame, and the first intersection points of the two perpendicular sides and the cancer cell boundary line, and taking the length of the cancer cell boundary line closest to the frame edge between the two first intersection points as the virtual length of the frame edge; If not, enter step S3; S3. Obtain the cancer cell data of each cancer cell at any moment, and calculate the similarity between the cancer cell data at the current moment and the cancer cell rectangular frames of the cancer cell data in history; S4. Set a similarity threshold, and determine whether the similarity value is greater than the similarity threshold; If so, extract the cancer cell number of the cancer cell data in history with the largest similarity value, and assign it to the cancer cell data at the current moment; If not, assign a new cancer cell number to the cancer cell data at the current moment; S5. Output monitoring data according to the cancer cell data, cancer cell numbers, and preset statistical rules.

2. The monitoring method according to claim 1, characterized in that, In step S2, the specific steps for obtaining the cancer cell data are as follows: S21. Establish an initial rectangular frame for each cancer cell according to the target detection algorithm, and obtain the cancer cell boundary line; S22. Select any one side of the rectangular frame as the active side, and mark the remaining sides as the passive sides; S23. Make the active side tangent to the cancer cell boundary line to obtain the tangent point, and make the passive side intersect with the cancer cell boundary line and have only one intersection point; S24. Move the tangent point along the cancer cell boundary line, and obtain the rectangular frame with the smallest area and with only one intersection point between the active side and the cancer cell boundary line, and mark it as the cancer cell rectangular frame; S25. Sequentially obtain two mutually perpendicular sides of each frame edge of the cancer cell rectangular frame, and the first intersection points of the two perpendicular sides and the cancer cell boundary line; S26. Obtain the length of the cancer cell boundary line closest to the frame edge between the two first intersection points, and use it as the virtual length of the frame edge; S27. Construct an expression of the cancer cell rectangular frame according to the frame edge size, angle, position of the cancer cell rectangular frame, and the virtual length of each frame edge to obtain the cancer cell data; the expression of the cancer cell rectangular frame is as follows: ； In the above formula, and respectively represent the first marker and the second marker of the cancer cell rectangle, and respectively represent the abscissa and ordinate of the center point of the cancer cell rectangle, and respectively represent the width and length of the cancer cell rectangle, where corresponds to the longest side in the cancer cell rectangle, S is the dataset of the virtual lengths of the cancer cell boundary lines corresponding to each frame side, represents the minimum angle passed when the X-axis rotates to be parallel to the side corresponding to any data in this dataset, .

3. The monitoring method according to claim 2, wherein In step S24, it specifically includes the following steps: S241. Randomly set a number of sampling points on the cancer cell boundary line, and move the tangent point along the boundary line for one week to obtain a number of first rectangular frames when the tangent point is located at the sampling points; S242. Screen out the first rectangular frames with two or more intersection points between the active side and the cancer cell boundary line, and mark them as the second rectangular frames; S243. Rotate the active side of the second rectangular frame and determine whether a rectangular frame can be obtained in which there is only one intersection point between the active side and the boundary line of the cancer cells; If so, rotate the active side of the second rectangular frame to obtain a new rectangular frame with the smallest area, and mark it as the first rectangular frame; If not, eliminate this second rectangular frame and the corresponding sampling point; S244. Select the first rectangular frame with the smallest area and in which there is only one intersection point between the active side and the boundary line of the cancer cells, and calculate the absolute value of the area difference between it and the previously obtained first rectangular frame; S245. Determine whether the absolute value of the area difference is less than the area difference threshold; If so, mark the current first rectangular frame as the cancer cell rectangular frame, and then proceed to step S25; If not, proceed to the next step; S246. Select the sampling point corresponding to the first rectangular frame with the smallest area, and mark the two sampling points adjacent to this sampling point as segmentation points; S247. Set several sampling points again between the two segmentation points, and move the tangent point along the boundary line of the cancer cells between the two segmentation points, and then return to step S242.

4. The monitoring method according to claim 3, wherein In step S245, the area difference threshold is 1.0% of the area of the previously obtained first rectangular frame with the smallest area.

5. The monitoring method according to claim 3, characterized in that, In step S241, the steps for setting the sampling points are as follows: S2411. Randomly set several initial sampling points on the boundary line of the cancer cells; S2412. Obtain each convex hull on the boundary line of the cancer cells; S2413. Calculate the ratio of the depth to the arc length of each convex hull in turn; S2414. Set the depth-to-arc-length ratio threshold, and eliminate the convex hulls whose ratio of depth to arc length is less than the depth-to-arc-length ratio threshold to obtain standard convex hulls; S2415. Set sampling points at the vertices, starting points, and ending points of the standard convex hulls.

6. The monitoring method according to claim 5, characterized in that, The number of the initial sampling points is greater than or equal to 12.

7. The monitoring method according to claim 1, characterized in that, In step S3, it specifically includes the following steps: S31. Rotate the angles of the cancer cell rectangular frames of the current moment and the cancer cell data in history around their center points to the same angle, and this angle is a multiple of 90°; S32. Calculate the IOU values between the cancer cell data at the current moment after conversion and the cancer cell rectangular frames corresponding to the cancer cell data in history; The calculation formula for the IOU value is: ； In the above formula, A and B respectively represent the cancer cell rectangular frames corresponding to the cancer cell data at the current moment and any cancer cell data in history. and respectively represent the abscissa and ordinate of the center point of the cancer cell rectangular frame A. and respectively represent the width and length of the cancer cell rectangular frame A. and respectively represent the abscissa and ordinate of the center point of the cancer cell rectangular frame B. and respectively represent the width and length of the object B. and respectively represent the horizontal and vertical overlap degrees of the cancer cell rectangular frame A and the cancer cell rectangular frame B. represents the overlapping area of the cancer cell rectangular frame A and the cancer cell rectangular frame B. and respectively represent the areas of the cancer cell rectangular frame A and the cancer cell rectangular frame B. represents the IOU value of the cancer cell rectangular frame A and the cancer cell rectangular frame B. S33. Calculate the similarity between the cancer cell data at the current moment and any cancer cell data in history according to the IOU value and the cancer cell data; The calculation formula for the similarity is: ； In the above formula, represents the similarity, and respectively represent the first coefficient and the second coefficient, represents the standard deviation of the difference between the data in the dataset of the virtual length in the cancer cell rectangle A and the corresponding data in the dataset of the virtual length in the cancer cell rectangle B.

8. The monitoring method according to claim 1, characterized in that In step S4, it specifically includes the following steps: S41. Set the similarity threshold, and determine whether the similarity value is greater than the similarity threshold; If so, extract the cancer cell number of the cancer cell data in history with the largest similarity value, and mark it as the candidate number of the current cancer cell data, and then proceed to the next step; If not, proceed to step S43; S42. Determine whether the cancer cell number that is the same as the candidate number of the current cancer cell data appeared at the previous moment; If so, use this candidate number as the cancer cell number of the current cancer cell data, and then proceed to step S5; If not, proceed to the next step; S43. Determine whether there is an intersection point between the cancer cell border of the current cancer cell data and other cancer cell data; If so, mark the cancer cell data as cancer cell data to be classified, and do not assign a cancer cell number to the cancer cell data, and then proceed to step S5; If not, proceed to the next step; S44. Determine whether the similarity value between the current cancer cell data and the cancer cell data in history is greater than the similarity threshold; If so, extract the cancer cell number of the cancer cell data in history with the largest similarity value, and mark it as the candidate number of the current cancer cell data, and then proceed to the next step; If not, assign a new cancer cell number to the cancer cell data at the current moment, and then proceed to the next step; S45. Starting from the current moment, calculate the similarity values between the cancer cell data at each moment and the cancer cell rectangle corresponding to the cancer cell data to be classified at the previous moment in turn; S46. Determine whether the similarity value is greater than the similarity threshold; If so, extract the cancer cell number of the cancer cell data with the largest similarity value, and assign it to the cancer cell data to be classified corresponding to the previous moment, and then proceed to the next step; If not, proceed to the next step.

9. The monitoring method according to claim 8, characterized in that, The cancer cell number is a positive integer, and each time a new cancer cell number is assigned to the cancer cell data at the current moment, the new cancer cell number is 1 greater than the largest cancer cell number value that has been generated.

10. A system for implementing the monitoring method of inhibiting the proliferation state of cervical cancer cells by the traditional Chinese medicine according to any one of claims 1-9, characterized in that, It includes a processor and a memory. The memory is used to store a computer program. When the computer program is executed by the processor, it implements the monitoring method for the inhibitory state of traditional Chinese medicine on the reproduction of cervical cancer cells according to any one of claims 1-9.

Images