Method and system for establishing mouse cerebral infarction model

Abstract

The invention relates to the technical field of animal experiments, and discloses a mouse cerebral infarction model establishment method and system.The mouse cerebral infarction model establishment method comprises the following steps that in the mouse cerebral infarction model operation process, mouse operation data are obtained in real time, and anomaly detection feedback is carried out, specifically, the anomaly detection feedback comprises generation of a suture insertion anomaly signal; if the suture insertion abnormal signal is generated, performing similarity analysis of curve shapes and characteristics in combination with a historical database of the mouse cerebral infarction model surgery, and constructing a significant similar curve set; and analyzing the effectiveness of the mouse cerebral infarction model operation in the significant similar curve set, and determining the adjustability of the current mouse cerebral infarction operation abnormality. According to the method, the pressure and depth data are collected in real time, the anomaly detection coefficient is calculated, immediate early warning of the suture insertion anomaly is carried out, whether adjustment is needed or an operation is ended is analyzed after early warning, the subjectivity of traditional experience judgment is reduced, and the success rate of the operation is increased.

Description

Technical Field

[0001] The present invention relates to the technical field of animal experiments, and particularly relates to a method and system for establishing a mouse cerebral infarction model. Background Art

[0002] The mouse cerebral infarction model is an important tool for simulating human ischemic stroke, and is widely used in the research and development of neuroprotective drugs, the study of angiogenesis mechanisms, and the evaluation of medical devices. Its core advantage is that by precisely controlling the ischemia time and location, focal cerebral infarction can be reproducibly induced. For example, the suture method can simulate the pathological process of thrombotic stroke by blocking the middle cerebral artery (MCAO). Its infarct volume and degree of neurological deficit are highly similar to human diseases.

[0003] However, the reliability of the mouse cerebral infarction model highly depends on the accuracy of surgical operation. The traditional suture method relies on the subjective experience of the operator to judge the insertion depth, lacking a quantitative feedback mechanism, which leads to a relatively high operation error and an increased coefficient of variation of the infarct volume. For example, when using a 10-0 nylon suture (diameter 0.25 mm), it is necessary to judge the position by touching the carotid artery pulsation, but it is easily affected by vasospasm or anatomical variation. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and system for establishing a mouse cerebral infarction model to solve at least one of the above-mentioned problems of the prior art.

[0005] In a first aspect, the present invention provides a method for establishing a mouse cerebral infarction model, including the following steps:

[0006] During the operation of establishing a mouse cerebral infarction model, real-time mouse operation data is acquired, and abnormal detection and feedback are performed. The abnormal detection and feedback include generating a suture insertion abnormal signal.

[0007] If a suture insertion abnormal signal is generated, in combination with the historical database of mouse cerebral infarction model operations, similarity analysis of curve shapes and features is performed to construct a set of significantly similar curves.

[0008] The effectiveness of the mouse cerebral infarction model operation in the set of significantly similar curves is analyzed to determine the adjustability of the current abnormality of the mouse cerebral infarction operation. If it is adjustable, the suture insertion depth of the mouse cerebral infarction model operation is adjusted.

[0009] As a further solution of the present invention: the process of performing abnormal detection and feedback is as follows:

[0010] Real-time data of the mouse operation is acquired, and a real-time pressure-depth curve is fitted. The real-time data includes: pressure value data and insertion depth value data.

[0011] Analyze and process real-time data to obtain an anomaly detection coefficient;

[0012] If the anomaly detection coefficient is greater than the anomaly detection coefficient threshold, generate a line plug insertion anomaly signal.

[0013] As a further solution of the present invention: the process of obtaining the anomaly detection coefficient is as follows:

[0014] Based on the real-time pressure-depth curve, calculate the slope change of adjacent curve segments;

[0015] Dynamically calculate the pressure mean and pressure standard deviation within the sliding window, and for the real-time pressure, calculate the standard score;

[0016] Calculate the difference between the real-time pressure value and the theoretical pressure value corresponding to the same depth, take the absolute value of the calculated difference, and then calculate the ratio with the pressure deviation limit value to obtain the pressure deviation ratio;

[0017] Based on the slope change rate, standard score, and pressure deviation ratio, use the principal component analysis model to calculate the anomaly detection coefficient.

[0018] As a further solution of the present invention: the process of performing curve shape similarity analysis is as follows:

[0019] Obtain the real-time pressure-depth curve and several historical pressure-depth curves;

[0020] Construct a real-time pressure-depth curve set and each historical pressure-depth curve set respectively;

[0021] Construct a distance matrix, based on the distance matrix, construct an accumulation matrix, and use a recurrence formula to fill the accumulation matrix;

[0022] Output the DTW distance, extract the historical pressure-depth curves corresponding to less than the DTW distance threshold, mark them as shape-similar curves, and construct a shape-similar curve set.

[0023] As a further solution of the present invention: the process of performing curve feature similarity analysis is as follows:

[0024] Analyze the shape-similar curve set to obtain a significant similarity index;

[0025] Extract the shape-similar curves greater than the significant similarity index threshold, mark them as significantly similar curves, and construct a significantly similar curve set.

[0026] As a further solution of the present invention: the process of obtaining the significant similarity index is as follows:

[0027] Obtain the slope change rate, standard score, and pressure deviation ratio of each shape-similar curve;

[0028] For the combination of a real-time pressure-depth curve and any one of the shape-similar curves;

[0029] Integrate the corresponding slope change rate sequence, standard score sequence, and pressure deviation ratio sequence into a feature vector, and calculate the cosine similarity;

[0030] Convert the DTW distance to a distance similarity;

[0031] Based on the distance similarity and cosine similarity, use the principal component analysis model to calculate and obtain the significant similarity index.

[0032] As a further solution of the present invention: The process of analyzing the effectiveness of the operation on the mouse cerebral infarction model in the set of significantly similar curves is as follows:

[0033] Based on the set of significantly similar curves, obtain the coefficient of variation of the infarct volume of the corresponding mouse model;

[0034] If the coefficient of variation of the infarct volume is less than the threshold of the coefficient of variation of the infarct volume, it is marked as an effective curve.

[0035] As a further solution of the present invention: The process of determining the adjustability of the abnormality of the current mouse cerebral infarction operation is as follows:

[0036] Count the number of effective curves, and perform a ratio calculation with the number of significantly similar curves to obtain the ratio of effective curves;

[0037] If the ratio of effective curves is greater than the limit of the ratio of effective curves, generate an adjustability signal.

[0038] As a further solution of the present invention: The process of adjusting the thread embolization insertion depth of the mouse cerebral infarction model operation is as follows:

[0039] Obtain the thread embolization insertion depth corresponding to the generation of the thread embolization insertion abnormal signal, denoted as the insertion depth abnormal value sdy;

[0040] Extract all effective curves and the corresponding insertion depth values, and calculate the average insertion depth sdj;

[0041] The adjusted insertion depth value sdt, the calculation formula is: , where α is an adjustment coefficient.

[0042] In a second aspect, the present invention provides a system for establishing a mouse cerebral infarction model, and the system includes:

[0043] Abnormal feedback module: During the operation of the mouse cerebral infarction model, real-time obtain the mouse operation data and perform abnormal detection and feedback, and the abnormal detection and feedback includes generating a thread embolization insertion abnormal signal;

[0044] Similarity analysis module: If an abnormal signal of wire embolization insertion is generated, in combination with the historical database of the mouse cerebral infarction model surgery, perform similarity analysis on the curve shape and characteristics to construct a set of significantly similar curves;

[0045] Adjustability judgment module: Analyze the effectiveness of the mouse cerebral infarction model surgery in the set of significantly similar curves to determine the adjustability of the current abnormality in the mouse cerebral infarction surgery;

[0046] Depth adjustment module: If it has adjustability, adjust the insertion depth of the wire embolization in the mouse cerebral infarction model surgery.

[0047] Advantages of the present invention:

[0048] 1. The present invention performs real-time acquisition of pressure and depth data and calculates the abnormal detection coefficient to give an immediate warning of abnormal wire embolization insertion. After the warning, it analyzes whether adjustment or termination of the surgery is required, reducing the subjectivity of traditional empirical judgment and increasing the success rate of the surgery;

[0049] 2. The present invention quickly excludes obviously mismatched curves from the shape level through the dynamic time warping algorithm, and further screens through the cosine similarity of feature vectors, improving the screening efficiency and accuracy. It adjusts the current surgery using the average insertion depth of historical effective cases, reducing the fluctuation of the infarction volume caused by vascular anatomical variation or operation error, and improving the rationality and safety of the establishment of the mouse model. Description of the drawings

[0050] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0051] Figure 1 is a flowchart of a method for establishing a mouse cerebral infarction model of the present invention;

[0052] Figure 2 is an architecture diagram of a system for establishing a mouse cerebral infarction model of the present invention. Detailed implementation manners

[0053] In order to enable those skilled in the art to better understand the solution of the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0054] Example 1

[0055] As Figure 1 shown, a method for establishing a mouse cerebral infarction model provided by an embodiment of the present invention specifically includes the following steps:

[0056] Step 1: During the operation of establishing a mouse cerebral infarction model, the operation data of the mouse are obtained in real time, and abnormal detection feedback is performed;

[0057] In some embodiments, during the operation of mouse cerebral infarction surgery, the real-time data of the mouse surgery are obtained, wherein the real-time data include: pressure value data and insertion depth value data;

[0058] Among them, the pressure value data is obtained through a wire tip integrated with a MEMS (Micro-Electro-Mechanical System) pressure sensor (such as a Kulite micro sensor);

[0059] The insertion depth value data is obtained through a micro displacement sensor. A micro displacement sensor (such as a strain gauge, a fiber Bragg grating sensor) is integrated at the tip or the tail of the wire, and the depth of the wire entering the blood vessel is measured in real time through mechanical transmission or optical signals;

[0060] Taking the insertion depth value and the pressure value at the same time as a set of data, fitting the real-time pressure-depth curve in a plane coordinate system, wherein the insertion depth value is the abscissa and the pressure value is the ordinate;

[0061] Calculate the slope change rate of adjacent curve segments , and its calculation formula is: ;

[0062] Among them, represents the slope value of the curve segment between the jth pressure data point and the (j - 1)th pressure data point, represents the slope value of the curve segment between the (j - 1)th pressure data point and the (j - 2)th pressure data point;

[0063] Dynamically calculate the pressure mean J and the pressure standard deviation B within the sliding window. Among them, the sliding window can be set by those skilled in the art based on the depth interval or the time interval, and at least 3 data points are required to calculate the standard deviation. The sliding window can be set to contain 5 - 20 data points to balance noise smoothing and real-time performance;

[0064] For the real-time pressure P, calculate the standard score Z, and the calculation formula is: ;

[0065] It should be noted that if the standard deviation is zero, it means that all the pressure data points included in the sliding window are the same, and there is no pressure change;

[0066] Dynamically compare the real-time pressure-depth curve with the theoretical pressure-depth curve;

[0067] Calculate the difference between the real-time pressure value and the theoretical pressure value corresponding to the same depth, take the absolute value of the calculated difference, and then calculate the ratio with the pressure deviation limit value to obtain the pressure deviation ratio;

[0068] Among them, the pressure deviation limit value is the maximum acceptable difference between the real-time pressure value and the theoretical pressure value at the same depth;

[0069] Among them, the process of constructing the theoretical pressure-depth curve is as follows:

[0070] Preoperative MRI data acquisition: Use a 7T small animal MRI system with an angiography sequence, set the resolution to 0.1 mm³ voxel, and the slice thickness to 0.2 mm;

[0071] Manually outline the contours of the common carotid artery (CCA), internal carotid artery (ICA), and middle cerebral artery (MCA) using ITK-SNAP software, and generate a three-dimensional vascular skeleton through image processing software;

[0072] The image processing software can be: the Skeletonize plugin of ImageJ;

[0073] Three-dimensional vascular parameter extraction: Equally divide the ICA-MCA path into several nodes, and obtain the coordinate positions of each node;

[0074] Among them, the nodes can be equally divided at intervals of 0.1 mm;

[0075] Fit an ellipse along the vascular cross-section at each node, and take the average of the major axis and the minor axis as the vascular diameter;

[0076] Calculate the tangent angle θ of three adjacent nodes, and the calculation formula is:

[0077] ;

[0078] Among them, θ is the included angle between the tangent vectors of node i and node i + 1 among three adjacent nodes, and is used to evaluate the curvature of the vascular path; represents the tangent vector of node i, indicating the direction of the vascular path at this node; represents the tangent vector of node i + 1, indicating the direction of the vascular path at the next node; represents the vector of the modulus, reflecting the magnitude of the tangent vector; represents the vector of the modulus, reflecting the magnitude of the tangent vector of the next node;

[0079] The included angle θ of the tangent reflects the degree of blood vessel curvature. Curved blood vessels will increase blood flow turbulence, thereby increasing resistance;

[0080] Segmented calculation of blood vessel resistance: Resistance of a single blood vessel segment Calculation formula:

[0081] ;

[0082] Among them, represents blood viscosity, with a value of 3.5 mPa·s, represents the length between nodes, with a value of the node interval, represents the radius, represents the diameter of a single blood vessel segment;

[0083] Pressure change of each segment , and the calculation formula is: , where the blood flow rate Q is 0.1 mL / min;

[0084] Theoretical pressure-depth curve fitting: Starting from the CCA bifurcation, accumulate the pressure segment by segment to obtain , and the accumulation formula is:

[0085] , among which, , n represents the total number of nodes;

[0086] Apply a filter to eliminate high-frequency noise, and use each group as coordinate points to draw the theoretical pressure-depth curve in sequence according to the depth order;

[0087] Based on the slope change rate, standard score, and pressure deviation ratio, use the principal component analysis model to calculate the anomaly detection coefficient. The specific process is as follows:

[0088] Obtain the three indicators of the slope change rate, standard score, and pressure deviation ratio corresponding to each real-time pressure data point, and organize them into a matrix;

[0089] Standardize the obtained data to obtain a standardized matrix;

[0090] Calculate the covariance matrix of the standardized matrix. Among them, the covariance matrix is a symmetric matrix;

[0091] Perform eigenvalue decomposition on the covariance matrix to obtain the eigenvalues corresponding to each indicator and the corresponding eigenvectors. The eigenvalue represents the variance of each principal component, and arrange the eigenvalues in descending order;

[0092] Calculate the contribution rate and cumulative contribution rate of each eigenvalue, select the principal components with a cumulative contribution rate greater than the contribution rate threshold, and construct a principal component matrix;

[0093] Multiply the standardized matrix by the principal component matrix to obtain the principal component score matrix;

[0094] According to the contribution rate of each principal component, perform a weighted sum of the principal component scores to obtain the anomaly detection coefficient;

[0095] The rate of change of the slope in the anomaly detection coefficient focuses on judging the situation of sudden pressure changes, the standard score focuses on eliminating individual differences and noise interference, and the standardized anomaly judgment and pressure deviation ratio focus on judging whether the suture deviates from the preset blood vessel path, such as the situation of misinserting into branches or abnormal resistance at the blood vessel bending point. The anomaly detection coefficient combines these three independent indicators and judges whether there is an abnormal suture insertion depth during the operation from different directions;

[0096] The reasons for judging whether the suture insertion depth is abnormal by analyzing the pressure data are as follows: one is that there are differences in the vascular anatomical structures of different mice, and the same insertion depth may correspond to different vascular positions; the other is that the insertion depth is the absolute value of mechanical displacement, but the pressure change can reflect the physiological state;

[0097] Set the anomaly detection coefficient threshold. If the anomaly detection coefficient is greater than the anomaly detection coefficient threshold, generate a suture insertion abnormal signal. If the anomaly detection coefficient is less than or equal to the anomaly detection coefficient threshold, generate a suture insertion normal signal;

[0098] The technical solution of this embodiment is: during the operation of the mouse cerebral infarction model, the pressure value and depth value during the suture insertion process are collected in real time, the pressure-depth curve is fitted, and the rate of change of the slope between adjacent curve segments, the standard score within the sliding window, and the deviation ratio of the real-time pressure to the theoretical pressure are calculated to comprehensively obtain the anomaly detection coefficient. If this coefficient exceeds the threshold, a suture insertion abnormal signal is generated;

[0099] Thus, through multi-dimensional quantitative analysis of the real-time operation data, dynamic monitoring and early warning of abnormal insertion can be realized, the accuracy of the operation is improved, and the problems of large operation errors and high coefficient of variation of the infarction volume caused by the traditional suture method relying on the operator's experience and lacking a quantitative feedback mechanism are avoided.

[0100] Embodiment 2

[0101] Based on the above embodiment, as Figure 1 shown, a method for establishing a mouse cerebral infarction model provided by an embodiment of the present invention specifically includes the following steps:

[0102] Step 2: Based on the suture insertion abnormal signal, combine the historical database of the mouse cerebral infarction model operation, perform curve similarity analysis, and construct a set of significantly similar curves;

[0103] In some embodiments, obtain the real-time pressure-depth curve and several historical pressure-depth curves;

[0104] Use the dynamic time warping (DTW) algorithm to obtain a set of shape-similar curves;

[0105] Preprocess the real-time pressure-depth curve and the historical pressure-depth curve respectively, where the preprocessing includes normalizing the insertion depth value and the pressure value;

[0106] Construct a set of real-time pressure-depth curves: , where, , represents the j-th real-time insertion depth value, represents the j-th real-time pressure value, and M represents the number of data groups of real-time insertion depth values and pressure values, ;

[0107] Construct a set of each historical pressure-depth curve respectively: , where, , represents the p-th historical insertion depth value, represents the p-th historical pressure value, and N represents the number of data groups of historical insertion depth values and pressure values, ;

[0108] Based on the set of real-time pressure-depth curves and any set of historical pressure-depth curves, construct an M×N distance matrix D;

[0109] where D[j, p] represents the Euclidean distance between the j-th data point of the real-time pressure-depth curve and the p-th data point of the historical pressure-depth curve , and the calculation formula is:

[0110] ;

[0111] Construct an M×N cumulative matrix A, where A[j, p] represents the minimum cumulative distance from the starting point (0, 0) to the point (j, p);

[0112] Set the initial conditions to initialize the cumulative matrix A:

[0113] ;

[0114] Use the recurrence formula to fill the cumulative matrix A, and the recurrence formula is:

[0115] ;

[0116] Backtrack the path starting from A[M - 1, N - 1]. The backtracking rule is: select the point corresponding to the minimum value among the predecessor points A[j - 1, p], A[j, p - 1], and A[j - 1, p - 1] of A[j, p], and record the path point (j, p) until returning to the starting point (0, 0);

[0117] Output the DTW distance as A[M - 1, N - 1], and the set of path points as the optimal matching point pairs;

[0118] The DTW distance quantifies the overall shape difference between the real - time pressure - depth curve and the historical pressure - depth curve. The smaller the value, the higher the similarity between the two curves;

[0119] Set the DTW distance threshold, extract the historical pressure - depth curves corresponding to less than the DTW distance threshold, mark them as shape - similar curves, and construct a set of shape - similar curves;

[0120] Based on the set of shape - similar curves, screen out the significantly similar curves;

[0121] Obtain the slope change rate, standard score, and pressure deviation ratio corresponding to each data point of each shape - similar curve,

[0122] Integrate the slope change rate sequence, standard score sequence, and pressure deviation ratio sequence corresponding to the real - time pressure - depth curve into a feature vector , and perform the same integration for any one of the shape - similar curves to obtain a feature vector ;

[0123] Based on the feature vectors of the real - time pressure - depth curve and any one of the shape - similar curves, calculate the cosine similarity S. The calculation formula is:

[0124] ;

[0125] where, is the dot product of vectors, is the vector norm;

[0126] Based on the DTW distance between the real - time pressure - depth curve and any one of the shape - similar curves , and convert it to a distance similarity . The conversion formula is:

[0127] ;

[0128] where, the larger the value of the distance similarity , the higher the similarity;

[0129] Based on the distance similarity and the cosine similarity, use the principal component analysis model to calculate the significantly similar index;

[0130] Among them, the calculation process is the same as that for calculating the abnormal detection coefficient;

[0131] Set the threshold of the significant similarity index, extract the shape-similar curves greater than the threshold of the significant similarity index, mark them as significant similarity curves, and construct a set of significant similarity curves;

[0132] It should be noted that if the set of curve similarity curves or the set of significant similarity curves is an empty set, it is determined that there is no matching historical case and the surgical abnormality cannot be adjusted. Those skilled in the art can choose:

[0133] Terminate the current surgery and check the experimental equipment, such as the calibration of the pressure sensor and the adaptability of the suture size;

[0134] Evaluate whether there are extreme variations in the blood vessels of the mouse, and you can choose to replace the mouse;

[0135] Mark the current surgical data as an unsuccessful case and store it to provide support for data analysis of future surgeries;

[0136] The advantage of first screening by the shape of the curve and then combining with feature screening is that there are many pressure-depth curves in the surgical history database. First, screening from the shape level by DTW can quickly exclude curves with obvious morphological mismatches. Curves with similar shapes may contain different types of abnormalities, such as vasospasm or too deep insertion of the suture. By analyzing the feature indicators through cosine similarity, it can be further determined whether the abnormal features are consistent;

[0137] Using a hierarchical and progressive screening logic increases the screening efficiency and accuracy, making the selected curves similar not only in the overall change trend of the curves but also in surgical features;

[0138] Step 3: Based on the set of significant similarity curves, perform an effectiveness analysis of the mouse cerebral infarction model surgery to determine the adjustability of the current mouse cerebral infarction surgery abnormality;

[0139] In some embodiments, based on the set of significant similarity curves, obtain the coefficient of variation of the infarction volume of the corresponding mouse model,

[0140] Compare the coefficient of variation of the infarction volume with the threshold of the coefficient of variation of the infarction volume. Among them, the threshold of the coefficient of variation of the infarction volume is set by those skilled in the art based on a comprehensive consideration of statistical power, experimental reproducibility, and industry standards, and can be set to 15%;

[0141] If the coefficient of variation of the infarction volume is less than the threshold of the coefficient of variation of the infarction volume, it indicates that the corresponding mouse infarction surgery has a high consistency, is a successful case, and is marked as an effective curve;

[0142] If the coefficient of variation of the infarct volume is greater than or equal to the threshold value of the coefficient of variation of the infarct volume, it indicates that the surgical effect of the corresponding mouse infarction is unreliable, which is a failed case and is marked as an invalid curve;

[0143] Count the number of valid curves and calculate the ratio with the number of significantly similar curves to obtain the ratio of valid curves;

[0144] Compare the ratio of valid curves with the limit value of the ratio of valid curves. Here, the limit value of the ratio of valid curves is used to judge the adjustability of the current mouse cerebral infarction model;

[0145] If the ratio of valid curves is greater than the limit value of the ratio of valid curves, it indicates that there are more successful cases in the historical surgeries similar to the current mouse cerebral infarction model, and an adjustability signal is generated;

[0146] If the ratio of valid curves is less than or equal to the limit value of the ratio of valid curves, it indicates that there are more failed cases in the historical surgeries similar to the current mouse cerebral infarction model, and a non-adjustability signal is generated;

[0147] Based on the non-adjustability signal, the surgery needs to be terminated or the mouse needs to be replaced;

[0148] Step Four: Based on the adjustability, adjust the insertion depth of the suture in the mouse cerebral infarction model surgery;

[0149] In some embodiments, obtain the insertion depth corresponding to the generation of the suture insertion abnormal signal, denoted as the abnormal insertion depth value sdy;

[0150] Extract all valid curves and the corresponding insertion depth values, and calculate the average insertion depth sdj;

[0151] It should be noted that the suture needs to be inserted along the internal carotid artery (ICA) into the middle cerebral artery (MCA). There is a physiological bend in the vascular path (such as the siphon segment of the internal carotid artery), and the insertion process is not strictly linearly increasing;

[0152] Then the adjusted insertion depth value sdt has the following calculation formula: , where α is the adjustment coefficient;

[0153] The adjustment coefficient is set by those skilled in the art according to the progress of the mouse cerebral infarction model surgery. Among them, it can be set to 1.0 in the early stage of the surgery (entering the common carotid artery), 0.7 in the middle stage of the surgery (entering the internal carotid artery), and 0.3 in the late stage of the surgery (entering the middle cerebral artery);

[0154] In addition, when adjusting the insertion depth of the suture, those skilled in the art also set corresponding adjustment range limit values for different periods of the surgery to avoid affecting the success rate of the surgery due to excessive adjustment;

[0155] The adjustment range limit is set based on the summary of multiple historical surgical experiments.

[0156] The technical solution of this embodiment is as follows: When an abnormal insertion of the suture is detected, the system performs shape matching on the real-time pressure-depth curve and historical data through the Dynamic Time Warping (DTW) algorithm, screens out a set of curves with similar shapes, further analyzes the cosine similarity of eigenvectors to construct a set of significantly similar curves. Based on the historical surgical effectiveness evaluation, if the proportion of effective cases meets the standard, the average insertion depth of the effective curves is extracted, and the insertion depth of the suture in the current surgery is dynamically corrected in combination with the adjustment coefficient.

[0157] Thus, through the DTW algorithm, obviously mismatched curves can be quickly excluded from the shape level, and further screened through the cosine similarity of eigenvectors, improving the screening efficiency and accuracy; adjusting the current surgery using the average insertion depth of historical effective cases to reduce the fluctuation of the infarct volume caused by vascular anatomical variations or operation errors; setting different adjustment coefficients according to the surgical stage to reduce the over-adjustment's impact on the surgical success rate and improve the rationality and safety of the adjustment plan.

[0158] Embodiment III

[0159] Based on the above embodiments, as Figure 2 shown, a system for establishing a mouse cerebral infarction model provided by an embodiment of the present invention specifically includes:

[0160] Abnormal feedback module: During the operation of establishing a mouse cerebral infarction model, the real-time surgical data of the mouse is acquired and abnormal detection feedback is performed.

[0161] In this embodiment, during the operation of mouse cerebral infarction surgery, the real-time data of the mouse surgery is acquired, where the real-time data includes: pressure value data and insertion depth value data.

[0162] Taking the insertion depth value and pressure value at the same time as a set of data, fitting the real-time pressure-depth curve in the plane coordinate system, where the insertion depth value is the abscissa and the pressure value is the ordinate.

[0163] Calculate the slope change rate of adjacent curve segments; dynamically calculate the pressure mean and pressure standard deviation within the sliding window; calculate the standard score for the real-time pressure.

[0164] Dynamically compare the real-time pressure-depth curve with the theoretical pressure-depth curve.

[0165] Calculate the difference between the real-time pressure value and the theoretical pressure value corresponding to the same depth, take the absolute value of the calculated difference, and then calculate the ratio with the pressure deviation limit to obtain the pressure deviation ratio.

[0166] Based on the slope change rate, standard score, and pressure deviation ratio, an anomaly detection coefficient is calculated using a principal component analysis model;

[0167] Set an anomaly detection coefficient threshold. If the anomaly detection coefficient is greater than the anomaly detection coefficient threshold, a wire plug insertion anomaly signal is generated. If the anomaly detection coefficient is less than or equal to the anomaly detection coefficient threshold, a wire plug insertion normal signal is generated;

[0168] Similarity analysis module: Based on the wire plug insertion anomaly signal, combined with the historical database of mouse cerebral infarction model surgery, perform curve similarity analysis to construct a set of significantly similar curves;

[0169] In this embodiment, a real-time pressure-depth curve and several historical pressure-depth curves are obtained;

[0170] Use the dynamic time warping (DTW) algorithm to obtain a set of shape-similar curves;

[0171] Preprocess the real-time pressure-depth curve and the historical pressure-depth curves respectively;

[0172] Construct a real-time pressure-depth curve set; construct each historical pressure-depth curve set respectively;

[0173] Based on the real-time pressure-depth curve set and any one historical pressure-depth curve set, construct an M×N distance matrix D;

[0174] Where D[j, p] represents the Euclidean distance between the j-th data point of the real-time pressure-depth curve and the p-th data point of the historical pressure-depth curve;

[0175] Construct an M×N cumulative matrix A; set initial conditions to initialize the cumulative matrix A; use a recurrence formula to fill the cumulative matrix A; backtrack the path starting from A[M - 1, N - 1];

[0176] Output the DTW distance as A[M - 1, N - 1], and the set of path points as the optimal matching point pairs;

[0177] Set a DTW distance threshold, extract the historical pressure-depth curves corresponding to less than the DTW distance threshold, mark them as shape-similar curves, and construct a set of shape-similar curves;

[0178] Based on the set of shape-similar curves, screen for significantly similar curves;

[0179] Obtain the slope change rate, standard score, and pressure deviation ratio corresponding to each data point of each shape-similar curve,

[0180] Integrate the slope change rate sequence, standard score sequence, and pressure deviation ratio sequence corresponding to the real-time pressure-depth curve into a feature vector. For any curve with a similar shape, perform the same integration to obtain a feature vector;

[0181] Calculate the cosine similarity based on the feature vectors of the real-time pressure-depth curve and any curve with a similar shape;

[0182] Based on the DTW distance between the real-time pressure-depth curve and any curve with a similar shape, convert it into a distance similarity;

[0183] Based on the distance similarity and cosine similarity, use the principal component analysis model to calculate and obtain a significant similarity index;

[0184] Set a threshold for the significant similarity index, extract the curves with a similar shape whose significant similarity index is greater than the threshold, mark them as significantly similar curves, and construct a set of significantly similar curves;

[0185] Adjustability judgment module: Based on the set of significantly similar curves, perform an effectiveness analysis of the mouse cerebral infarction model surgery to determine the adjustability of the abnormality of the current mouse cerebral infarction surgery;

[0186] In this embodiment, based on the set of significantly similar curves, obtain the coefficient of variation of the infarction volume of the corresponding mouse model;

[0187] Compare the coefficient of variation of the infarction volume with the threshold of the coefficient of variation of the infarction volume;

[0188] If the coefficient of variation of the infarction volume is less than the threshold of the coefficient of variation of the infarction volume, it indicates that the infarction surgery of the corresponding mouse has a high consistency, which is a successful case and is marked as an effective curve;

[0189] If the coefficient of variation of the infarction volume is greater than or equal to the threshold of the coefficient of variation of the infarction volume, it indicates that the effect of the infarction surgery of the corresponding mouse is unreliable, which is a failed case and is marked as an invalid curve;

[0190] Count the number of effective curves, and calculate the ratio of the number of effective curves to the number of significantly similar curves to obtain the ratio of effective curves;

[0191] Compare the ratio of effective curves with the limit ratio of effective curves;

[0192] If the ratio of effective curves is greater than the limit ratio of effective curves, it indicates that there are more successful cases in the historical surgeries similar to the current mouse cerebral infarction model, and generate an adjustability signal;

[0193] Depth adjustment module: Based on the adjustability, adjust the insertion depth of the suture in the mouse cerebral infarction model surgery;

[0194] In this embodiment, the plug insertion depth corresponding to the abnormal signal of plug insertion is obtained and denoted as the abnormal value of insertion depth sdy;

[0195] All valid curves and the corresponding insertion depth values are extracted, and the average insertion depth sdj is calculated;

[0196] Then, the adjusted insertion depth value sdt is calculated by the following formula: , where α is the adjustment coefficient.

[0197] In the above embodiments, the descriptions of the various embodiments have their own emphases. For the parts not detailed or recorded in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0198] The above has described in detail one embodiment of the present invention, but the content described is only a preferred embodiment of the present invention and cannot be considered as limiting the scope of implementation of the present invention. All equivalent changes and improvements made according to the scope of the application of the present invention should still fall within the scope covered by the patent of the present invention.

Claims

1. A method for establishing a mouse cerebral infarction model, characterized in that It includes the following steps: During the operation of the mouse cerebral infarction model, real-time mouse operation data is obtained and anomaly detection feedback is performed. The anomaly detection feedback includes generating a signal for abnormal suture insertion; If a signal for abnormal suture insertion is generated, similarity analysis of curve shapes and features is performed in combination with the historical database of the mouse cerebral infarction model operation to construct a set of significantly similar curves; For the effectiveness analysis of the mouse cerebral infarction model operation in the set of significantly similar curves, the adjustability of the current anomaly in the mouse cerebral infarction operation is determined. If it is adjustable, the insertion depth of the suture in the mouse cerebral infarction model operation is adjusted.

2. The method for establishing a mouse cerebral infarction model according to claim 1, wherein, The process of performing anomaly detection feedback is as follows: Real-time data of the mouse operation is obtained, and a real-time pressure-depth curve is fitted. The real-time data includes pressure value data and insertion depth value data; The real-time data is analyzed and processed to obtain an anomaly detection coefficient. If the anomaly detection coefficient is greater than the anomaly detection coefficient threshold, a signal for abnormal suture insertion is generated.

3. The method for establishing a mouse cerebral infarction model according to claim 2, wherein The process of obtaining the anomaly detection coefficient is as follows: Based on the real-time pressure-depth curve, the slope change of adjacent curve segments is calculated; The pressure mean and pressure standard deviation within the sliding window are dynamically calculated. For the real-time pressure, the standard score is calculated; The difference between the real-time pressure value and the theoretical pressure value corresponding to the same depth is calculated, the absolute value of the calculated difference is taken, and then the ratio is calculated with the pressure deviation limit value to obtain the pressure deviation ratio; Based on the slope change rate, standard score, and pressure deviation ratio, using the principal component analysis model, the anomaly detection coefficient is calculated.

4. The method for establishing a mouse cerebral infarction model according to claim 1, characterized in that, The process of performing curve shape similarity analysis is as follows: The real-time pressure-depth curve and several historical pressure-depth curves are obtained; A real-time pressure-depth curve set and each historical pressure-depth curve set are respectively constructed; A distance matrix is constructed. Based on the distance matrix, a cumulative matrix is constructed, and the cumulative matrix is filled using a recurrence formula; The DTW distance is output, and the historical pressure-depth curves corresponding to less than the DTW distance threshold are extracted, marked as shape-similar curves, and a set of shape-similar curves is constructed.

5. The method for establishing a mouse cerebral infarction model according to claim 4, characterized in that, The process of performing curve feature similarity analysis is as follows: The set of shape-similar curves is analyzed to obtain a significant similarity index; The shape-similar curves greater than the significant similarity index threshold are extracted, marked as significantly similar curves, and a set of significantly similar curves is constructed.

6. The method for establishing a mouse cerebral infarction model according to claim 5, wherein, The process of obtaining the significant similarity index is as follows: The slope change rate, standard score, and pressure deviation ratio of each shape-similar curve are obtained; For the combination of the real-time pressure-depth curve and any one shape-similar curve; The corresponding slope change rate sequence, standard score sequence, and pressure deviation ratio sequence are integrated into a feature vector, and the cosine similarity is calculated; The DTW distance is converted into a distance similarity; Based on the distance similarity and cosine similarity, using the principal component analysis model, the significant similarity index is calculated.

7. The method for establishing a mouse cerebral infarction model according to claim 1, wherein The process of performing effectiveness analysis of the mouse cerebral infarction model operation in the set of significantly similar curves is as follows: Based on the set of significantly similar curves, the coefficient of variation of the infarction volume of the corresponding mouse model is obtained; If the coefficient of variation of the infarction volume is less than the coefficient of variation threshold of the infarction volume, it is marked as an effective curve.

8. The method for establishing a mouse cerebral infarction model according to claim 7, wherein The process of determining the adjustability of the current mouse cerebral infarction surgery abnormality is as follows: Count the number of valid curves, calculate the ratio with the number of significantly similar curves, and obtain the ratio of valid curves; If the ratio of valid curves is greater than the limit value of the ratio of valid curves, generate an adjustability signal.

9. The method for establishing a mouse cerebral infarction model according to claim 1, characterized in that, The process of adjusting the insertion depth of the suture in the mouse cerebral infarction model surgery is as follows: Obtain the insertion depth corresponding to the generated suture insertion abnormality signal, denoted as the abnormal insertion depth value sdy, extract all valid curves and the corresponding insertion depth values, and calculate the average insertion depth sdj; The adjusted insertion depth value sdt, the calculation formula is: , where α is the adjustment coefficient.

10. A system for establishing a mouse cerebral infarction model, characterized in that, This system is used to execute the method described in any one of claims 1-9 above. This system includes: Abnormal feedback module: During the process of performing mouse cerebral infarction model surgery, real-time obtain mouse surgery data and perform abnormal detection and feedback. The abnormal detection and feedback include generating a suture insertion abnormality signal; Similarity analysis module: If a suture insertion abnormality signal is generated, combine the historical database of mouse cerebral infarction model surgery to perform similarity analysis of curve shapes and characteristics, and construct a set of significantly similar curves; Adjustability judgment module: Analyze the effectiveness of the mouse cerebral infarction model surgery in the set of significantly similar curves to determine the adjustability of the current mouse cerebral infarction surgery abnormality; Depth adjustment module: If it is adjustable, adjust the insertion depth of the suture in the mouse cerebral infarction model surgery.

Images