Local red blood cell aggregation degree determination method and system based on ultrasonic detection

Through ultrasonic detection technology, the blood flow rate zone and the red blood cell aggregation degree is calculated, which solves the problem of insufficient detection of red blood cell aggregation degree in the existing technology, and the accurate evaluation of the red blood cell aggregation status is achieved, and it has important clinical diagnostic value.

CN119027405APending Publication Date: 2024-11-26YUNNAN UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411196918.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-29
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

The existing red blood cell aggregation detection technology cannot accurately simulate the red blood cell aggregation status in real situations, and does not consider the impact of blood flow velocity shear rate on red blood cell aggregation, resulting in inaccurate detection results.

Method used

Using ultrasound detection method, the ultrasound image of the human carotid artery is obtained, divided into high-flow and low-flow areas of blood. The speckle tracking method and area growth algorithm are used to identify red blood cell aggregates, image features are extracted and red blood cell aggregation value is calculated. Consider the effect of blood flow velocity on red blood cell aggregation.

Benefits of technology

It improves the accuracy of detection of red blood cell aggregation and can better reflect the aggregation of red blood cell in the human body, which is of great significance especially in early disease diagnosis.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119027405A_ABST
    Figure CN119027405A_ABST
Patent Text Reader

Abstract

The invention discloses a local erythrocyte aggregation degree determination method and system based on ultrasonic detection, and relates to the field of medical image detection.The method comprises the steps that a speckle tracking method is adopted to track blood vessel wall internal area images of a current frame and a next frame to obtain a blood flow velocity curve, and a blood flow velocity threshold value is determined based on the blood flow velocity curve; cutting the current frame image into a blood high-flow-rate area image and a blood low-flow-rate area image based on the blood flow velocity curve and a threshold value; red blood cell aggregates on the blood flow velocity region image are identified based on a region growing algorithm, the segmented blood flow velocity region image is obtained, feature extraction is carried out, and image features are obtained; and based on the image features, calculating an erythrocyte aggregation degree value in the blood flow velocity region image by adopting an erythrocyte aggregation degree calculation formula. According to the method, the influence of the blood flow velocity shear rate on the erythrocyte aggregation degree is considered, the real erythrocyte aggregation state is simulated, and the accuracy of erythrocyte aggregation detection is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of medical image detection, and in particular to a method and system for determining local red blood cell aggregation based on ultrasonic detection. Background Art

[0002] In the blood circulation system, when the shear force is low, red blood cells will attract each other to form a coin-like string or larger two-dimensional or three-dimensional irregular aggregates. Under normal circumstances, the aggregates of red blood cells in the human body can be properly dispersed by blood flow, but in various pathological processes such as metabolic disorders, cardiovascular and cerebrovascular diseases, diabetes, and infection, this aggregation effect is enhanced, and even causes complications such as thrombosis. Related studies have shown that the increase in the degree of red blood cell aggregation in patients with angina pectoris significantly increases the risk of myocardial infarction. The probability of myocardial infarction in patients with angina pectoris with highly aggregated red blood cells is 5.7 times that of patients with angina pectoris whose red blood cell aggregation remains within the normal range. Other researchers believe that increased red blood cell aggregation is independently associated with stroke, that is, abnormal red blood cell aggregation is a precursor to stroke. Therefore, red blood cell aggregation can be used as an indicator for early detection and auxiliary diagnosis of related diseases, and its graded analysis has important clinical significance.

[0003] In order to achieve the classification of red blood cell aggregation, some researchers have divided the red blood cell aggregation into a limited number of categories and adopted traditional classifiers based on data features or directly used deep neural networks of the data itself to achieve the classification of red blood cell aggregation. However, the assumption that red blood cell aggregation is divided into a limited number of categories is far from the actual red blood cell aggregation. For example, the red blood cell aggregation is divided into three types: low aggregation, moderate aggregation, and high aggregation. The classification interval is large and cannot simulate the real red blood cell aggregation. Furthermore, the blood flow velocity in the blood vessel is parabolic, that is, the blood flow velocity is highest near the central axis of the blood vessel, and the blood flow velocity is smaller the closer to the blood vessel wall. The above method does not consider the effect of blood flow velocity shear rate on red blood cell aggregation. Summary of the invention

[0004] The purpose of the present application is to provide a method and system for determining local red blood cell aggregation based on ultrasonic detection, so as to solve the problem that the existing red blood cell aggregation division interval is too large to simulate the real red blood cell aggregation state and does not consider the influence of blood flow velocity shear rate on the red blood cell aggregation degree.

[0005] To achieve the above objectives, this application provides the following solutions:

[0006] In a first aspect, the present application provides a method for determining local red blood cell aggregation based on ultrasound detection, comprising:

[0007] Acquire a human carotid artery ultrasound image, and based on the human carotid artery ultrasound image, obtain a plurality of continuous frames of blood vessel wall internal area images, and determine any frame of the blood vessel wall internal area image as a current frame image, and determine the next frame of the blood vessel wall internal area image of the current frame image as a next frame image;

[0008] Using a speckle tracking method to track the current frame image and the next frame image to obtain a blood flow velocity curve, and determining a blood flow velocity threshold based on the blood flow velocity curve;

[0009] Based on the blood flow velocity curve and the blood flow velocity threshold, cutting the current frame image into a high blood flow velocity area image and a low blood flow velocity area image;

[0010] Based on the region growing algorithm, the red blood cell aggregates on the blood flow area image are identified, and the identified red blood cell aggregates are set to white, and the area outside the red blood cell aggregates is set to black, so as to obtain the segmented blood flow area image; the blood flow area image includes a blood high flow area image and a blood low flow area image;

[0011] Extracting features from the segmented blood flow rate area image to obtain image features; the image features include the number of white independent connected domains, the area of ​​the white independent connected domains, the perimeter of the white independent connected domains, the area of ​​the circumscribed ellipse of the white independent connected domains, the area of ​​the black area, and the total area of ​​the segmented blood flow rate area image;

[0012] Based on the image features, the red blood cell aggregation value in the blood flow rate area image is calculated using a red blood cell aggregation calculation formula.

[0013] In a second aspect, the present application provides a system for determining local red blood cell aggregation based on ultrasound detection, wherein the system for determining local red blood cell aggregation based on ultrasound detection is based on the method for determining local red blood cell aggregation based on ultrasound detection, and the system for determining local red blood cell aggregation based on ultrasound detection comprises:

[0014] An image determination unit is used to obtain an ultrasonic image of a human carotid artery, and based on the ultrasonic image of the human carotid artery, obtain a plurality of continuous frames of images of the inner region of the blood vessel wall, and determine any frame of the inner region of the blood vessel wall as a current frame image, and determine the next frame of the inner region of the blood vessel wall of the current frame image as a next frame image;

[0015] A blood flow velocity curve and blood flow velocity threshold determination unit, configured to track the current frame image and the next frame image using a speckle tracking method to obtain a blood flow velocity curve, and determine a blood flow velocity threshold based on the blood flow velocity curve;

[0016] A blood flow velocity region image cropping unit, configured to crop a current frame image into a blood high flow velocity region image and a blood low flow velocity region image based on the blood flow velocity curve and the blood flow velocity threshold;

[0017] The segmented blood flow velocity area image determination unit is used to identify the red blood cell aggregates on the blood flow velocity area image based on the region growing algorithm, and set the identified red blood cell aggregates to white, and set the area outside the red blood cell aggregates to black, so as to obtain the segmented blood flow velocity area image; the blood flow velocity area image includes a blood high flow velocity area image and a blood low flow velocity area image;

[0018] A feature extraction unit is used to extract features from the segmented blood flow velocity area image to obtain image features; the image features include the number of white independent connected domains, the area of ​​the white independent connected domains, the perimeter of the white independent connected domains, the area of ​​the circumscribed ellipse of the white independent connected domains, the area of ​​the black area, and the total area of ​​the segmented blood flow velocity area image;

[0019] The red blood cell aggregation value determination unit is used to calculate the red blood cell aggregation value in the blood flow rate area image based on the image features and using the red blood cell aggregation calculation formula.

[0020] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0021] The present application provides a method and system for determining local red blood cell aggregation based on ultrasonic detection. By dividing an ultrasonic image into a high blood flow rate area and a low blood flow rate area, red blood cell aggregation values ​​in different flow rate areas can be obtained, thereby solving the problem that the existing red blood cell aggregation detection technology does not consider the influence of blood flow rate on red blood cell aggregation. The present application considers the influence of blood flow rate on red blood cell aggregation, thereby making the detection result more accurate. Based on image features, a red blood cell aggregation calculation formula is used to calculate the red blood cell aggregation value in the blood flow rate area image, and the specific numerical value of the red blood cell aggregation is directly calculated. This solves the problem that the existing red blood cell aggregation detection technology has too large a division interval between red blood cell aggregation classes and cannot simulate the real red blood cell aggregation situation of the human body, thereby improving the detection accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0023] Figure 1This is a diagram of the application environment of a method for determining local red blood cell aggregation based on ultrasonic detection in one embodiment of the present application;

[0024] Figure 2 A schematic diagram of a flow chart of a method for determining local red blood cell aggregation based on ultrasonic detection provided in one embodiment of the present application;

[0025] Figure 3 A schematic diagram of defining a tracking block in an embodiment of the present application;

[0026] Figure 4 A schematic diagram of determining the best matching block in an embodiment of the present application;

[0027] Figure 5 A schematic diagram of calculating the distance between the center point coordinates of the tracking block and the center point coordinates of the best matching block in an embodiment of the present application;

[0028] Figure 6 A schematic diagram of determining a partition node tracking block in an embodiment of the present application;

[0029] Figure 7 This is a schematic diagram of dividing a high blood flow rate area and a low blood flow rate area in one embodiment of the present application;

[0030] Figure 8 Schematic diagram of cropped blood high flow rate area image and blood low flow rate area image in one embodiment of the present application;

[0031] Fig. 9 Another schematic diagram of a flow chart of a method for determining local red blood cell aggregation based on ultrasound detection provided by another embodiment of the present application;

[0032] Fig.10 A schematic diagram of the functional modules of a system for determining local red blood cell aggregation based on ultrasound detection provided in another embodiment of the present application.

[0033] Reference numerals:

[0034] Image determination unit-1, blood flow velocity curve and blood flow velocity threshold determination unit-2, blood flow velocity area image clipping unit-3, segmented blood flow velocity area image determination unit-4, feature extraction unit-5, red blood cell aggregation value determination unit-6. DETAILED DESCRIPTION

[0035] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0036] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0037] The method for determining local red blood cell aggregation based on ultrasound detection provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the human carotid artery ultrasound image to the server 104. After the server 104 receives the human carotid artery ultrasound image, the server 104 processes the human carotid artery ultrasound image and calculates the red blood cell aggregation value in the blood flow rate area image. The server 104 can feed back the obtained red blood cell aggregation value to the terminal 102. In addition, in some embodiments, the local red blood cell aggregation determination method based on ultrasonic detection can also be implemented separately by the server 104 or the terminal 102, such as the terminal 102 can directly determine the local red blood cell aggregation based on ultrasonic detection for the human carotid artery ultrasound image, or the server 104 can obtain the human carotid artery ultrasound image from the data storage system, and determine the local red blood cell aggregation based on ultrasonic detection for the human carotid artery ultrasound image.

[0038] The terminal 102 may be, but is not limited to, various desktop computers, laptop computers, smart phones, tablet computers, IoT devices, and portable wearable devices. The IoT devices may be smart speakers, smart TVs, smart air conditioners, smart vehicle-mounted devices, etc. The portable wearable devices may be smart watches, smart bracelets, head-mounted devices, etc. The server 104 may be implemented as an independent server or a server cluster consisting of multiple servers, or may be a cloud server.

[0039] In an exemplary embodiment, Figure 2 As shown, a method for determining local red blood cell aggregation based on ultrasonic detection is provided. The method is executed by a computer device, and can be executed by a computer device such as a terminal or a server alone, or by a terminal and a server together. In the embodiment of the present application, the method is applied to Figure 1 The server 104 in the example is used for explanation, and the steps include the following steps S1 to S6.

[0040] in:

[0041] Step S1, obtaining a human carotid artery ultrasound image, and based on the human carotid artery ultrasound image, obtaining multiple frames of continuous blood vessel wall internal area images, and determining any frame of the blood vessel wall internal area image as the current frame image, and determining the next frame of the blood vessel wall internal area image of the current frame image as the next frame image.

[0042] Step S2: Track the current frame image and the next frame image using a speckle tracking method to obtain a blood flow velocity curve, and determine a blood flow velocity threshold based on the blood flow velocity curve.

[0043] Step S3, based on the blood flow velocity curve and the blood flow velocity threshold, cutting the current frame image into a high blood flow velocity area image and a low blood flow velocity area image.

[0044] Step S4, based on the region growing algorithm, identify the red blood cell aggregates on the blood flow area image, set the identified red blood cell aggregates to white, and set the area outside the red blood cell aggregates to black, to obtain the segmented blood flow area image; the blood flow area image includes a high blood flow area image and a low blood flow area image.

[0045] Step S5, extracting features from the segmented blood flow velocity area image to obtain image features; the image features include the number of white independent connected domains, the area of ​​the white independent connected domains, the perimeter of the white independent connected domains, the area of ​​the circumscribed ellipse of the white independent connected domains, the area of ​​the black area, and the total area of ​​the segmented blood flow velocity area image.

[0046] Step S6, based on the image features, using the red blood cell aggregation calculation formula to calculate the red blood cell aggregation value in the blood flow rate area image.

[0047] By implementing the above steps S1 to S6, the red blood cell aggregation values ​​in different flow rate areas can be obtained, which solves the problem that the existing red blood cell aggregation detection technology does not consider the influence of blood flow rate on red blood cell aggregation, and the detection result is more accurate; based on image features, the red blood cell aggregation calculation formula is used to calculate the red blood cell aggregation value in the blood flow rate area image, which solves the problem that the existing red blood cell aggregation detection technology has too large a division interval between red blood cell aggregation classes and cannot simulate the real red blood cell aggregation situation of the human body.

[0048] In another exemplary embodiment of the present application, step S1 specifically includes:

[0049] Using ultrasound equipment to collect ultrasound images of human carotid arteries I o And store, with size l o × o The rectangular frame of I o Intercept and obtain the internal area image I of the blood vessel wall cropped , the width of the rectangle wo The principle of selection is: ensure that its upper and lower frames fit tightly to the inner side of the blood vessel wall.

[0050] In another exemplary embodiment of the present application, step S2 specifically includes:

[0051] Step S21, divide the current frame image into multiple transverse strip areas along the blood flow direction, use a window of a preset size located at the leftmost position of each transverse strip area as a tracking block, and determine the coordinates of the center point of the tracking block. The tracking block is defined as Figure 3 Specifically, the speckle tracking method is used to track the i-th frame image. and the i+1th frame image The blood flow speckle is Divided into N along the blood flow direction blocks horizontal strip area, take the leftmost strip area as The window is used as the tracking block B t ,Right now There are N blocks Tracking blocks, record N blocks The center point of the tracking block is The coordinates in are: (x track_j ,y track_k ), k=1,2,…,N blocks .

[0052] Step S22, for each tracking block, based on the coordinates of the center point of the tracking block, determine the corresponding area of ​​the tracking block in the next frame image, use the area outside the corresponding area as a search area, and define multiple matching blocks of the same size as the tracking block in the search area.

[0053] Specifically, in the i+1th frame image According to the kth (k=1,2,…,N blocks ) coordinates of the center point of the tracking block (x track_k ,y track_k ) locates the tracking block and defines a (l o -l b )×w s The search area is defined as l in the search area. b × b The matching block B m , there are N in the search area match (N match =(l o -l b )×w s ) matching blocks.

[0054] Step S23, respectively calculating the normalized correlation coefficients between the tracking block and each of the matching blocks, and determining the best matching block of the tracking block according to all the normalized correlation coefficients. The determined best matching block is as follows: Figure 4 shown.

[0055] Specifically, calculate the kth (k = 1, 2, ..., N blocks ) tracking blocks and their N match The normalized correlation coefficient NCC of the matching blocks is used to find the best matching block of the k-th tracking block, where the matching block corresponding to the minimum NCC value is the best matching block of the k-th tracking block. The expression of the normalized correlation coefficient is as follows:

[0056]

[0057] Among them, NCC k_q represents the normalized correlation coefficient between the kth tracking block and the qth matching block; B t_k (x i ,x j ) indicates that the coordinates of the kth tracking block are (x i ,x j )’s pixel value, x i =1,2,…,w b , x j =1,2,…,l b ; is the mean value of all pixels in the kth tracking block; B m_q (x i ,x j ) is the coordinate of the qth matching block (x i ,x j )’s pixel value, x i =1,2,…,w b , x j =1,2,…,l b ; is the mean of all pixels in the qth matching block.

[0058] Step S24, determine the center point coordinates of the best matching block, calculate the distance between the center point coordinates of the tracking block and the center point coordinates of the best matching block, the schematic diagram of the distance between the center point coordinates of the tracking block and the center point coordinates of the best matching block is as follows: Figure 5 shown.

[0059] Specifically, the center point coordinates of the best matching block are (x match_k ,y match_k ), k=1,2,…,N blocks , record the center point coordinates of the best matching blocks corresponding to all tracking blocks; the center point of the kth tracking block (x track_k ,ytrack_k ) from the center point of its best matching block (x match_k ,y match_k ) k The expression is:

[0060]

[0061] Among them, x match_k ,y match_k are the horizontal and vertical coordinates of the center point of the kth best matching block, respectively, x track_k ,y track_k are the horizontal and vertical coordinates of the center point of the kth tracking block respectively.

[0062] Step S25, determining the average motion speed of the tracking block between the current frame image and the next frame image based on the distance between the center point coordinates of the tracking block and the center point coordinates of the best matching block and the time interval between the current frame image and the next frame image.

[0063] Specifically, s k Divide the two frames The time interval t between the two frames is used to obtain the average motion speed v of the kth tracking block between two adjacent frames. k :

[0064]

[0065] Among them, v k is the motion speed of the kth tracking block, and t is The time interval between these two frames of images of the internal area of ​​the blood vessel wall.

[0066] Step S26, obtaining the blood flow velocity curve according to the average motion speed of the tracking block between the current frame image and the next frame image. blocks The average movement speed of the tracking blocks is obtained, that is, a blood flow velocity curve C is obtained, and the maximum velocity in C is recorded as v max .

[0067] In another exemplary embodiment of the present application, step S3 specifically includes:

[0068] Step S31, selecting a split node tracking block based on the blood flow velocity curve and the blood flow velocity threshold; the split node tracking block is two tracking blocks whose average motion speed is equal to the blood flow velocity threshold, such as Figure 6 shown.

[0069] like Figure 6As shown, on the blood flow velocity curve C, find the divided node tracking blocks n1 and n2 corresponding to the motion speed equal to the blood flow velocity threshold θ, that is, calculate the horizontal coordinates x of the two points where the blood flow velocity curve C intersects with the horizontal line v=θ θ1 、x θ2 , if x θ1 、x θ2 is an integer, then n1=x θ1 , n2=x θ2 ; if x θ1 、x θ2 If it is not an integer, take the value from x θ1 、x θ2 The nearest integer value is assigned to n1 and n2.

[0070] Step S32, determine two horizontal lines passing through the center coordinate point of the division node tracking block, use the two horizontal lines as division lines, divide the area between the two horizontal lines into a high blood flow rate area, and divide the remaining area into a low blood flow rate area, such as Figure 7 shown.

[0071] Specifically, along the two horizontal lines passing through the center points of the n1 and n2 tracking blocks, Divide and crop into blood high flow rate area image I h Image I of the low blood flow area l_1 ,I l_2 Among them, the calculation formula of blood flow velocity threshold θ is:

[0072]

[0073] Step S33, cropping the image corresponding to the high blood flow rate area to obtain the high blood flow rate area image; cropping the image corresponding to the low blood flow rate area to obtain the low blood flow rate area image. The divided high blood flow rate area image and low blood flow rate area image are shown in FIG. Figure 8 shown.

[0074] In another exemplary embodiment of the present application, step S4 specifically includes:

[0075] Step S41, preprocessing the blood flow velocity area image, and using a connected component analysis method to identify independent connected domains in the preprocessed blood flow velocity area image, and calculating the centroid coordinates of each independent connected domain.

[0076] Step S42, determining a preset range in the preprocessed blood flow rate area image, taking the area within the preset range as the area to be detected, and taking the centroid coordinates as the seed growth point coordinates corresponding to the area to be detected.

[0077] Step S43: using the seed growth point as the starting calculation point of the region growing algorithm in the current region to be detected, and performing region growing detection on the current region to be detected.

[0078] Step S44, taking the area formed by the seed growth points set as white in each area to be detected as the identified red blood cell aggregate.

[0079] In another exemplary embodiment of the present application, step S41 specifically includes:

[0080] Step S411, defining an edge preserving operator; the edge preserving operator includes a horizontal convolution kernel and a vertical convolution kernel. The horizontal convolution kernel and the vertical convolution kernel E of the defined edge preserving operator x With E y Respectively expressed as:

[0081]

[0082] Step S412, determining a first target matrix of each pixel point in the blood flow velocity region image, and taking a 3×3 matrix centered at a certain pixel point (x, y) in the blood flow velocity region image as the first target matrix.

[0083] Step S413, convolve the first target matrix of each pixel with the horizontal convolution kernel and the vertical convolution kernel to obtain the horizontal edge gradient detection value and the vertical edge gradient detection value. Specifically, the first target matrix is ​​convoluted with the horizontal convolution kernel E of the edge preservation operator (Edge_preserve operator) x And the vertical convolution kernel E y Perform convolution operation to obtain the horizontal edge gradient detection value G x and vertical edge gradient detection value G y .

[0084] Step S414, based on the horizontal edge gradient detection value and the vertical edge gradient detection value, determine the edge gradient calculation value of each pixel point. The expression for calculating the edge gradient calculation value G of the pixel point (x, y) is as follows:

[0085]

[0086] Step S415: determine the gradient matrix based on the edge gradient calculation value of each pixel. Traverse all the pixels in the blood flow rate area image to calculate their edge gradient calculation value G, and combine the G values ​​of all the pixels into a gradient matrix M. gradient .

[0087] Step S416, performing a convolution operation on the gradient matrix and an edge-preserving denoising filter to filter out weak echoes in the blood flow rate area image, thereby obtaining a numerical matrix.

[0088] Specifically, M gradient Convolution operation is performed with the self-designed edge-preserving filter to filter out weak echoes, and the numerical matrix M is obtained. n :

[0089]

[0090] Among them, s i ,d i It is half the size of the edge-preserving filter kernel, and the filter kernel is 3×3 in size, so s i =1,d i = 1. K(s,d) is the value of the sth row and dth column in the kernel matrix K of the edge-preserving filter. The kernel matrix K is:

[0091]

[0092] Step S417, performing a convolution operation on the numerical matrix and a Gaussian filter to filter out noise in the blood flow velocity area image, thereby obtaining a filtered blood flow velocity area image.

[0093] Specifically, M n Convolution operation is performed with a Gaussian filter to filter out image noise and obtain the filtered blood flow velocity area image I filtered :

[0094]

[0095] Among them, the kernel function of the Gaussian filter is:

[0096]

[0097] Here, σ is the standard deviation of the Gaussian function.

[0098] Step S418, binarization processing is performed on the filtered blood flow velocity area image to obtain a pre-processed blood flow velocity area image. Specifically:

[0099] a. The filtered blood flow area image I filtered Convert to grayscale image I gray :

[0100] I gray =0.2989·x f_red +0.5870 x f_green +0.1140 x f_blue .

[0101] Among them, xf_red 、x f_green 、x f_blue I filtered The RGB value of the f-th pixel in the image.

[0102] b. Normalize the grayscale image to obtain the normalized blood flow rate area image I norm :

[0103]

[0104] Among them, x g Represents the grayscale image I gray The value of the g-th pixel in the g ) is the grayscale image I gray The minimum pixel value in g ) is the grayscale image I gray The maximum pixel value in .

[0105] c. Image binarization based on local adaptive method to obtain the binary blood flow velocity area image I binary :

[0106] Traverse the normalized blood flow velocity area image I norm All pixels in the calculation are based on pixel x n_i The binarization threshold ε of the matrix M (size l×c) centered at n_i :

[0107]

[0108] Among them, k s is a constant used to adjust the threshold sensitivity, n = l × c, is the i-th n Row j n The value of the column, μ is the mean of all numbers in the matrix M, and the calculation formula of μ is:

[0109]

[0110] d. Compare pixel x n_i The value of ε n_i If x n_i >ε n_i , then the pixel value of the pixel is set to white (pixel value: 255); if x n_i <ε n_i , then the pixel value of the pixel is set to black (pixel value: 0), the expression is:

[0111]

[0112] Traversal I normAll pixels in the image are obtained to obtain the preprocessed blood flow velocity area image I binary .

[0113] Step S419, select a second target matrix centered on the current pixel from the preprocessed blood flow rate area image, and divide the adjacent pixels with the same pixel value as the current pixel into the same independent connected domain as the current pixel; the adjacent pixels are the pixels in the second target matrix other than the current pixel.

[0114] Specifically, the binarized blood flow rate area image I binary Middle b i Line b j The pixel coordinates of the column are (b i ,b j ), in pixels (b i ,b j ) as the 3×3 matrix centered as the second target matrix, and the remaining pixels in the second target matrix as (b i ,b j ) of adjacent pixels; in (b i ,b j ) of adjacent pixels, the pixel value is the same as (b i ,b j ) is the same as (b i ,b j ) are recorded as pixels in the same connected domain, and the pixel values ​​are the same as (b i ,b j ) is different from (b i ,b j ) do not belong to the same connected domain and will not be recorded.

[0115] Step S420, traverse all the pixels in the preprocessed blood flow velocity region image to determine a plurality of independent connected domains. binary All pixels in the final N con Connected domains, where the pixel point record format contained in a single connected domain is:

[0116] [(z,e),…,(u,v),…,(o,r)].

[0117] Among them, (z,e), (u,v), and (o,r) represent the coordinates of pixels belonging to the same connected domain.

[0118] Step S421: Calculate the centroid coordinates of each independent connected domain based on the number of pixels contained in each independent connected domain and the corresponding pixel coordinates. con In the k-th connected domain, c (k c =1,2,…,Ncon ) connected domain centroid coordinates The calculation formula is:

[0119]

[0120] in, For the kth c The number of pixels in the k-th connected domain; z, u, o are c The horizontal coordinates of the pixels in the kth connected domain; e, v, r are c The vertical coordinates of the pixels in the connected domain; the binary blood flow velocity area image I binary The coordinate set S of all centroids in centroid The record format is:

[0121] The image is binarized and the coordinates of the centroid of the independent connected region are counted. The position of the red blood cell aggregates is determined by combining the filtered blood flow area image and the centroid coordinates of the independent connected domain, providing the coordinates of the seed growth points for the subsequent regional growing algorithm to obtain the segmented image. The image binarization process aims to lock the approximate distribution position of the red blood cell aggregates, laying the foundation for the subsequent refined image segmentation.

[0122] In another exemplary embodiment of the present application, step S43 specifically includes:

[0123] Define the seed growth point, S centroid N stored in con The centroid coordinates are taken as I filtered The coordinates of the seed growth point, in I filtered k s (k s =1,2,…,N con ) seed growth points As the kth s The starting point of the region growing algorithm for each region is I filtered There are N con The seed growth point is the first growth point to be detected in the area to be detected, and then the adjacent pixels of the growth point to be detected are used as new growth points for regional growth detection. This process is repeated until the adjacent pixels of the growth point set to black are all set to black, and the regional growth algorithm of the area stops. Specifically:

[0124] Step S431, obtaining the pixel grayscale value of the current seed growth point and the grayscale values ​​of the adjacent pixels of the current seed growth point, and calculating the regional differentiation threshold of the current seed growth point according to the pixel grayscale value of the current seed growth point and the grayscale values ​​of the adjacent pixels of the current seed growth point.

[0125] Among them, the expression for calculating the regional differentiation threshold of the current seed growth point is:

[0126]

[0127] In step S432, the expression for calculating the regional difference of the current seed growth point is:

[0128]

[0129] Among them, δ represents the regional differentiation threshold; I(x f ,y f ) indicates that the coordinate is (x f ,y f )’s gray value; p h The coordinates are (x f ,y f )’s grayscale value of the hth neighboring pixel; Differentiation indicates the regional difference; The coordinates are (x f ,y f ) and the mean grayscale value of its adjacent pixels.

[0130] Step S432, calculating the regional difference of the current seed growth point according to the pixel grayscale value of the current seed growth point and the average of the pixel grayscale values ​​of the current seed growth point and adjacent pixels.

[0131] Step S433: when the regional difference of the current seed growth point is less than or equal to the regional difference threshold of the current seed growth point, the current seed growth point is set to white.

[0132] Step S434: when the regional difference of the current seed growth point is greater than the regional difference threshold of the current seed growth point, the current seed growth point is set to black.

[0133] Step S435, performing region growth detection on all growth points in the current region to be detected until the adjacent pixels of the growth points set to black in the current region to be detected are all set to black, thereby terminating the region growth detection of the current region to be detected.

[0134] Preferably, after completing the region growth detection on the current region to be detected, the method further comprises: filtered The unassigned pixels in the image are set to black, and the segmented image I is obtained. seg .

[0135] In another exemplary embodiment of the present application, step S5 specifically includes:

[0136] Count and record segmented image I segThe number of white connected regions (representing red blood cell aggregates) in the middle is N, and the area of ​​white independent connected domains is S a , the perimeter of the white independent connected domain C s , and discard the white connected domains with an area less than 20 pixels, and calculate the ellipse area S of the white independent connected domain t , calculate the area S of the black area in the segmented image b , the total area of ​​the segmented blood flow velocity zone image is denoted as S.

[0137] In another exemplary embodiment of the present application, step S6 specifically includes:

[0138] The formula for calculating red blood cell aggregation is:

[0139]

[0140] Among them, R i represents the calculated aggregation value of the red blood cell aggregate corresponding to the i-th white independent connected domain; 0 <i≤N;C s represents the perimeter of the white independent connected domain; S a represents the area of ​​the white independent connected domain; S t represents the circumscribed ellipse area of ​​the white independent connected domain; R is the calculated value of the red blood cell aggregation in the current blood flow rate area image; N represents the number of white independent connected domains; It represents the normalized aggregation value of the red blood cell aggregate corresponding to the first white independent connected domain; represents the normalized aggregation value of the red blood cell aggregate corresponding to the Nth white independent connected domain; S represents the total area of ​​the segmented blood flow area image; S b Indicates the area of ​​the black region.

[0141] Beneficial effects of this application:

[0142] 1) By dividing the image of the internal area of ​​the blood vessel wall into a high blood flow rate area and a low blood flow rate area, the red blood cell aggregation is detected, and the influence of blood flow rate on the red blood cell aggregation is considered, so that the detection result is more accurate, and the red blood cell aggregation values ​​in different flow rate areas can be obtained, which solves the problem that the existing red blood cell aggregation detection technology does not consider the influence of blood flow rate on red blood cell aggregation.

[0143] 2) This application provides for the first time a calculation formula for red blood cell aggregation, which can be used to obtain specific red blood cell aggregation values. This solves the problem in existing red blood cell aggregation detection technologies that the red blood cell aggregation interval is divided too large and cannot simulate the real human red blood cell aggregation situation. It can reflect the subtle effects of related diseases in the early stages on red blood cell aggregation.

[0144] 3) The image of the internal area of ​​the blood vessel wall is binarized to determine the approximate location of the red blood cell aggregates, and then the region growing algorithm is used to finely segment the red blood cell aggregates. This unsupervised segmentation method is intended to avoid the influence of subjective judgment on the manually labeled data in the supervised method.

[0145] 4) The edge detection operator used by the filter is a custom operator, which is used to remove weak echo signals in ultrasound images to highlight the red blood cell aggregate information in the image.

[0146] The method for determining local red blood cell aggregation based on ultrasonic detection of the present application is further described in detail below in conjunction with another exemplary embodiment.

[0147] 1) Collecting human carotid artery ultrasound images I o and store.

[0148] 2) If Fig. 9 As shown, with a size of l o × o =237×154 rectangular frame interception I o Image of the inner area of ​​the blood vessel wall I cropped , using speckle tracking to track two adjacent frames I cropped The blood flow speckle of the image is used to obtain the blood flow velocity curve. According to the spatial position corresponding to the blood flow velocity threshold θ, I cropped Divide and crop into blood high flow rate area image I h Image I of the low blood flow area l The specific steps are:

[0149] a. Track the i-th frame using speckle tracking With the i+1th frame image The blood flow speckle is used to obtain the blood flow velocity curve:

[0150] ① The current frame image Divided into N along the blood flow direction blocks = 30 horizontal strips, the leftmost strip is The window is used as the tracking block B t ,Right now There are N blocks = 30 tracking blocks, record N blocks The center point of the tracking block is The coordinates in are: (x track_k ,y track_k ), k=1,2,…,30.

[0151] ② In the next frame image According to the kth (k=1,2,…,N blocks ) coordinates of the center point of the tracking block (xtrack_k ,y track_k ) locates the tracking block and defines a ( l o -l b )×w s =232×15 search area, and define the size of l in the search area b × b =5×5 matching block B m , there are N in the search area match (N match =(l o -l b )×w s =3480) matching blocks.

[0152] ③ Calculate the kth tracking block and its N match =The normalized correlation coefficient NCC of 3480 matching blocks, find the best matching block of the kth tracking block.

[0153] ④ Calculate N in sequence blocks The movement speed of the tracking block is obtained by a blood flow velocity curve C, and the maximum velocity in C is recorded as v max =0.77m / s.

[0154] b. On the blood flow velocity curve C, according to the blood flow velocity threshold Find the corresponding tracking block numbers n1 and n2, and get n1 = 8, n2 = 24. Follow the horizontal straight line passing through the center points of the n1th and n2th tracking blocks to Divide and crop into blood high flow rate area image I h Image I of the low blood flow area l_1 ,I l_2 ,like Fig. 9 shown.

[0155] c. Repeat steps a and b to save all I cropped Divided into blood high flow rate area image I h Image I of the low blood flow area l .

[0156] 3) Use edge-preserving denoising filter to remove the high blood flow rate area image I h Image I of the low blood flow area l_1 ,I l_2 A single high velocity area image I h For example:

[0157] a. Put I h All pixels in the horizontal convolution kernel E of the Edge_preserve operator xAnd the vertical convolution kernel E y Perform convolution operation to obtain the horizontal edge gradient detection value G x and vertical edge gradient detection value G y After that, the edge gradient calculation value G of all pixels is calculated, and the G values ​​of all pixels are combined into a gradient matrix M gradient .

[0158] b. M gradient Convolution operation is performed with the self-designed edge-preserving filter to filter out weak echoes, and the numerical matrix M is obtained. n .

[0159] c. M n Convolution operation is performed with a Gaussian filter to filter out image noise and obtain the filtered blood high flow rate area image I filtered .

[0160] 4) Will I filtered Copy Image binarization based on local adaptive method Convert it into a grayscale image and perform normalization and binarization operations in sequence to obtain a binary blood high flow rate area image I binary , where the image binarization based on the local adaptive method obtains the binary blood flow velocity area image I binary , binarization threshold ε i for:

[0161]

[0162] Among them, M is the pixel x n_i The matrix centered on is of size n=l×c=3×3, is the i-th n Row j n The value of the column, k s =1.23 is a constant used to adjust the threshold sensitivity, and μ is the mean of all numbers in the matrix M.

[0163] Compare pixels x n_i The value of ε n_i If x n_i >ε n_i , then the pixel is set to white (pixel value: 255); if x n_i <ε n_i , then the pixel is set to black (pixel value: 0), traversing I norm All pixels in the image are binarized to obtain the high blood flow rate area image I binary .

[0164] 5) Use connected component analysis to identify I binaryThe independent connected regions in the , calculate and store the centroid coordinates of a single connected region

[0165] 6) Use the region growing algorithm to identify the I obtained in step 3) filtered The red blood cell aggregates on the image are set to white, and the remaining areas are set to black to obtain the segmented image I seg The specific steps are:

[0166] a. Take all centroid coordinates obtained in step 5) as I filtered The coordinates of the seed growth point.

[0167] b. Take the seed growth point as the starting point for regional growth detection. The coordinates are (x f ,y f ) of the pixel value I(x f ,y f ) is the region growing criterion:

[0168]

[0169] Where Differentiation is the calculated point I(x f ,y f ), and δ is the calculated pixel difference threshold of the area.

[0170] c.To I filtered The unassigned pixels in the image are set to black, and the segmented image I is obtained. seg . Low flow rate area image I l_1 ,I l_2 and high velocity area image I h The corresponding segmented images I are obtained by image processing seg_l_1 ,I seg_l_2 ,I seg_h ,like Fig. 9 shown.

[0171] 7) Now Fig. 9 Segmentation image of medium and high flow rate area I seg_h For example, the image features can be extracted as follows:

[0172] The number of white independent connected domains: N=46.

[0173] White independent connected domain area: S a1 =21.5, S a2 =235, S a3 =58.5, S a4 =158.5, S a5 =2786.5, S a6 =430, S a7=20.5, S a8 =265,S a9 =971,S a10 =300, S a11 =164.5, S a12 =582.5, S a13 =168862,S a14 =98,S a15 =38.5, S a16 =417.5, S a17 =422, S a18 =272, S a19 =175, S a20 =65, S a21 =3070.5, S a22 =357, S a23 =33.5, S a24 =330.5, S a25 =396.5, S a26 =28, S a27 =31, S a28 =46.5, S a29 =927, S a30 =201,S a31 =488.5, S a32 =42.5, S a33 =116.5, S a34 =1033.5,S a35 =1209,S a36 =212,S a37 =88,S a38 =656, S a39 =452, S a40 =381, S a41 =1581.5,S a42 =33035,S a43 =391.5, S a44 =25.5, S a45 =273, S a46 =1558.5.

[0174] The perimeter of the white independent connected domain: C s1 =33.21, C s2 =67.46, C s3 =46.53, C s4 =51.7, C s5 =236.75, C s6 =119.05, C s7 =32.87, C s8 =80.43, C s9=153.68, C s10 =90.23, C s11 =102.12, C s12 =186.55, C s13 =7505.86, C s14 =63.8, C s15 =29.21, C s16 =148.47, C s17 =188.23, C s18 =154.43, C s19 =92.43, C s20 =42.97, C s21 =541.49, C s22 =84.43, C s23 =31.07, C s24 =142.33, C s25 =111.84, C s26 =24.97, C s27 =31.8, C s28 =29.21, C s29 =239.28, C s30 =80.08, C s31 =146.33, C s32 =32.04, C s33 =64.53, C s34 =243.1, C s35 =244.85, C s36 =111.46, C s37 =74.97, C s38 =240.85, C s39 =221.54, C s40 =164.08, C s41 =266.41, C s42 =1886.85, C s43 =140.12, C s44 =23.9, C s45 =126.08, C s46 =335.46.

[0175] The area of ​​the circumscribed ellipse of the white independent connected domain: S t1 =32.42, S t2 =256.72, S t3 =88.88, S t4 =169.22, S t5 =2994.5, S t6 =523.7, S t7 =34.54, St8 =283.38, S t9 =1005.74, S t10 =332.04, S t11 =247.78, S t12 =720.54, S t13 =201751.55,S t14 =104.8, S t15 =53.82, S t16 =447.72, S t17 =281.77, S t18 =370.74, S t19 =215.21, S t20 =82.51, S t21 =4440.97, S t22 =412.07, S t23 =39.03, S t24 =416.61, S t25 =455.72, S t26 =34.95, S t27 =47.29, S t28 =50.49, S t29 =1596.72, S t30 =250.38, S t31 =586.19, S t32 =52.53, S t33 =134.08, S t34 =2048.54, S t35 =1454.36, S t36 =221.67, S t37 =100.53, S t38 =1333.81, S t39 =733.22, S t40 =407.14, S t41 =1828.85, S t42 =44364.94, S t43 =489.58, S t44 =28.83, S t45 =339.98, S t46 =2619.01.

[0176] The total area of ​​the segmented high blood flow rate area image is: S=277830.

[0177] The area of ​​the black area in the segmented image: S b =54521.

[0178] 8) Segment the blood high flow rate area image I obtained in step 7) seg_h Substitute all the image feature values ​​into the formula for calculation to obtain the calculated aggregation value of a single aggregate:

[0179]

[0180] Compare all the calculated aggregation values ​​of individual aggregates, and record the largest calculated value as R max The smallest calculated value is recorded as R min . Normalize all calculated aggregation values ​​of individual aggregates to one:

[0181]

[0182] The blood high flow rate area image I in the specific embodiment is calculated h Calculated value of red blood cell aggregation:

[0183]

[0184] The blood high flow rate area image I can be obtained by calculation h The calculated value of red blood cell aggregation R h is 0.15. Similarly, the blood low flow rate area image I can be obtained l_1 The calculated value of red blood cell aggregation R l_1 is 0.2431, blood low flow rate area image I l_2 The calculated value of red blood cell aggregation R l_2 It is 0.2486.

[0185] Based on the same inventive concept, the embodiment of the present application also provides a system for determining the local red blood cell aggregation based on ultrasound detection for implementing the method for determining the local red blood cell aggregation based on ultrasound detection mentioned above. The implementation scheme for solving the problem provided by the system is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more embodiments of the system for determining the local red blood cell aggregation based on ultrasound detection provided below can refer to the limitations of the method for determining the local red blood cell aggregation based on ultrasound detection above, and will not be repeated here.

[0186] In an exemplary embodiment, Fig.10 As shown, a local red blood cell aggregation determination system based on ultrasonic detection is provided, including: an image determination unit-1, a blood flow velocity curve and a blood flow velocity threshold determination unit-2, a blood flow velocity area image cropping unit-3, a segmented blood flow velocity area image determination unit-4, a feature extraction unit-5, and a red blood cell aggregation value determination unit-6.

[0187] The image determination unit 1 is used to obtain an ultrasonic image of a human carotid artery and obtain an image of an internal area of ​​a blood vessel wall based on the ultrasonic image of the human carotid artery.

[0188] The blood flow velocity curve and blood flow velocity threshold determination unit 2 is used to track the internal area images of the blood vessel wall in two adjacent frames using a speckle tracking method to obtain a blood flow velocity curve, and determine a blood flow velocity threshold based on the blood flow velocity curve.

[0189] The blood flow velocity region image clipping unit 3 is used to clip the current frame blood vessel wall internal region image into a blood high flow velocity region image and a blood low flow velocity region image based on the blood flow velocity curve and the blood flow velocity threshold.

[0190] The segmented blood flow velocity area image determination unit 4 is used to identify the red blood cell aggregates on the blood flow velocity area image based on the region growing algorithm, and set the identified red blood cell aggregates to white, and set the area outside the red blood cell aggregates to black, so as to obtain the segmented blood flow velocity area image; the blood flow velocity area image includes a high blood flow velocity area image and a low blood flow velocity area image.

[0191] The feature extraction unit 5 is used to extract features from the segmented blood flow velocity area image to obtain image features; the image features include the number of white independent connected domains, the area of ​​the white independent connected domains, the perimeter of the white independent connected domains, the area of ​​the circumscribed ellipse of the white independent connected domains, the area of ​​the black area and the total area of ​​the segmented blood flow velocity area image.

[0192] The red blood cell aggregation value determination unit 6 is used to calculate the red blood cell aggregation value in the blood flow rate area image based on the image features and using the red blood cell aggregation calculation formula.

[0193] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0194] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0195] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0196] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0197] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A method for determining local red blood cell aggregation based on ultrasonic detection, characterized in that: The method for determining local red blood cell aggregation based on ultrasound detection includes: Acquire a human carotid artery ultrasound image, and based on the human carotid artery ultrasound image, obtain a plurality of continuous frames of blood vessel wall internal area images, and determine any frame of the blood vessel wall internal area image as a current frame image, and determine the next frame of the blood vessel wall internal area image of the current frame image as a next frame image; Using a speckle tracking method to track the current frame image and the next frame image to obtain a blood flow velocity curve, and determining a blood flow velocity threshold based on the blood flow velocity curve; Based on the blood flow velocity curve and the blood flow velocity threshold, cutting the current frame image into a high blood flow velocity area image and a low blood flow velocity area image; Based on the region growing algorithm, the red blood cell aggregates on the blood flow area image are identified, and the identified red blood cell aggregates are set to white, and the area outside the red blood cell aggregates is set to black, so as to obtain the segmented blood flow area image; the blood flow area image includes a blood high flow area image and a blood low flow area image; Extracting features from the segmented blood flow rate area image to obtain image features; the image features include the number of white independent connected domains, the area of ​​the white independent connected domains, the perimeter of the white independent connected domains, the area of ​​the circumscribed ellipse of the white independent connected domains, the area of ​​the black area, and the total area of ​​the segmented blood flow rate area image; Based on the image features, the red blood cell aggregation value in the blood flow rate area image is calculated using a red blood cell aggregation calculation formula.

2. The method for determining local red blood cell aggregation based on ultrasonic detection according to claim 1, characterized in that The speckle tracking method is used to track the current frame image and the next frame image to obtain the blood flow velocity curve, which specifically includes: The current frame image is divided into a plurality of transverse strip regions along the blood flow direction, a window of a preset size located at the leftmost position of each transverse strip region is used as a tracking block, and the coordinates of the center point of the tracking block are determined; For each tracking block, based on the coordinates of the center point of the tracking block, determine the corresponding area of ​​the tracking block in the next frame image, use the area outside the corresponding area as a search area, and define multiple matching blocks with the same size as the tracking block in the search area; respectively calculating normalized correlation coefficients between the tracking block and each of the matching blocks, and determining the best matching block of the tracking block according to all the normalized correlation coefficients; Determine the center point coordinates of the best matching block, and calculate the distance between the center point coordinates of the tracking block and the center point coordinates of the best matching block; Determine the average motion speed of the tracking block between the current frame image and the next frame image based on the distance between the center point coordinates of the tracking block and the center point coordinates of the best matching block and the time interval between the current frame image and the next frame image; The blood flow velocity curve is obtained according to the average movement speed of the tracking block between the current frame image and the next frame image.

3. The method for determining local red blood cell aggregation based on ultrasonic detection according to claim 1, characterized in that: Based on the blood flow velocity curve and the blood flow velocity threshold, the current frame image is cut into a blood high flow velocity area image and a blood low flow velocity area image, specifically including: Selecting a division node tracking block based on the blood flow velocity curve and the blood flow velocity threshold; the division node tracking blocks are two tracking blocks whose average motion speed is equal to the blood flow velocity threshold; Determine two horizontal straight lines passing through the center coordinate point of the division node tracking block, divide the area between the two horizontal straight lines into a high blood flow velocity area, and divide the remaining area into a low blood flow velocity area; The image corresponding to the high blood flow rate area is cropped to obtain the high blood flow rate area image; the image corresponding to the low blood flow rate area is cropped to obtain the low blood flow rate area image.

4. The method for determining local red blood cell aggregation based on ultrasonic detection according to claim 1, characterized in that: Based on the region growing algorithm, red blood cell aggregates on the blood flow area image are identified, including: Preprocessing the blood flow velocity region image, identifying independent connected domains in the preprocessed blood flow velocity region image using a connected component analysis method, and calculating the centroid coordinates of each independent connected domain; Determine a preset range in the preprocessed blood flow rate area image, use the area within the preset range as the area to be detected, and use the centroid coordinates as the seed growth point coordinates of the corresponding area to be detected; The seed growth point is used as the starting calculation point of the region growing algorithm in the current region to be detected, and the region growing detection is performed on the current region to be detected; The area formed by the seed growth points set as white in each area to be detected is regarded as the identified red blood cell aggregate.

5. The method for determining local red blood cell aggregation based on ultrasonic detection according to claim 4, characterized in that: Preprocessing the blood flow velocity area image specifically includes: An edge preserving operator is defined; the edge preserving operator includes a horizontal convolution kernel and a vertical convolution kernel; Determine a first target matrix of each pixel in the blood flow velocity area image; Convolution operations are performed on the first target matrix of each pixel point with the horizontal convolution kernel and the vertical convolution kernel respectively to obtain a horizontal edge gradient detection value and a vertical edge gradient detection value; Determine the edge gradient calculation value of each pixel based on the horizontal edge gradient detection value and the vertical edge gradient detection value; Determine the gradient matrix based on the edge gradient calculation value of each pixel point; Performing a convolution operation on the gradient matrix and an edge-preserving denoising filter to filter out weak echoes in the blood flow rate area image to obtain a numerical matrix; Performing a convolution operation on the numerical matrix and a Gaussian filter to filter out noise in the blood flow velocity area image, thereby obtaining a filtered blood flow velocity area image; The filtered blood flow velocity area image is binarized.

6. The method for determining local red blood cell aggregation based on ultrasonic detection according to claim 5, characterized in that: The connected component analysis method is used to identify independent connected domains in the preprocessed blood flow area image, and the centroid coordinates of each independent connected domain are calculated, specifically including: A second target matrix centered on the current pixel is selected from the preprocessed blood flow rate area image, and adjacent pixels having the same pixel value as the current pixel are divided into the same independent connected domain as the current pixel; the adjacent pixels are pixels other than the current pixel in the second target matrix; Traversing all pixel points in the preprocessed blood flow velocity area image to determine multiple independent connected domains; Based on the number of pixels contained in each independent connected domain and the corresponding coordinates of each pixel point, the centroid coordinates of each independent connected domain are calculated.

7. The method for determining local red blood cell aggregation based on ultrasonic detection according to claim 4, characterized in that: The seed growth point is used as the starting calculation point of the region growth algorithm in the current region to be detected, and the region growth detection is performed on the current region to be detected, specifically including: Obtaining the pixel grayscale value of the current seed growth point and the grayscale values ​​of the adjacent pixels of the current seed growth point, and calculating the regional differentiation threshold of the current seed growth point according to the pixel grayscale value of the current seed growth point and the grayscale values ​​of the adjacent pixels of the current seed growth point; Calculate the regional difference of the current seed growth point according to the pixel gray value of the current seed growth point and the average of the pixel gray values ​​of the current seed growth point and adjacent pixels; When the regional difference of the current seed growth point is less than or equal to the regional difference threshold of the current seed growth point, the current seed growth point is set to white; When the regional difference of the current seed growth point is greater than the regional difference threshold of the current seed growth point, the current seed growth point is set to black; The regional growth detection is performed on all the growth points in the current region to be detected until the adjacent pixels of the growth points set to black in the current region to be detected are all set to black, and the regional growth detection of the current region to be detected is terminated.

8. The method for determining local red blood cell aggregation based on ultrasonic detection according to claim 7, characterized in that: The expression for calculating the regional differentiation threshold of the current seed growth point is: The expression for calculating the regional difference of the current seed growth point is: Among them, δ represents the regional differentiation threshold; I(x f ,y f ) indicates that the coordinate is (x f ,y f )’s gray value; p h The coordinates are (x f ,y f )’s grayscale value of the hth neighboring pixel; Differentiation indicates the regional difference; The coordinates are (x f ,y f ) and the mean grayscale value of its adjacent pixels.

9. The method for determining local red blood cell aggregation based on ultrasonic detection according to claim 1, characterized in that: The expression for calculating the red blood cell aggregation value in the blood flow rate area image is: Among them, R i represents the calculated aggregation value of the red blood cell aggregate corresponding to the i-th white independent connected domain; C s represents the perimeter of the white independent connected domain; S a represents the area of ​​the white independent connected domain; S t represents the circumscribed ellipse area of ​​the white independent connected domain; R is the calculated value of the red blood cell aggregation in the current blood flow rate area image; N represents the number of white independent connected domains; It represents the normalized aggregation value of the red blood cell aggregate corresponding to the first white independent connected domain; represents the normalized aggregation value of the red blood cell aggregate corresponding to the Nth white independent connected domain; S represents the total area of ​​the segmented blood flow area image; S b Indicates the area of ​​the black region.

10. A system for determining local red blood cell aggregation based on ultrasonic detection, the system for determining local red blood cell aggregation based on ultrasonic detection is based on the method for determining local red blood cell aggregation based on ultrasonic detection according to any one of claims 1 to 9, characterized in that: The local red blood cell aggregation determination system based on ultrasound detection comprises: An image determination unit is used to obtain an ultrasonic image of a human carotid artery, and based on the ultrasonic image of the human carotid artery, obtain a plurality of continuous frames of images of the inner region of the blood vessel wall, and determine any frame of the inner region of the blood vessel wall as a current frame image, and determine the next frame of the inner region of the blood vessel wall of the current frame image as a next frame image; A blood flow velocity curve and blood flow velocity threshold determination unit, configured to track the current frame image and the next frame image using a speckle tracking method to obtain a blood flow velocity curve, and determine a blood flow velocity threshold based on the blood flow velocity curve; A blood flow velocity region image cropping unit, configured to crop a current frame image into a blood high flow velocity region image and a blood low flow velocity region image based on the blood flow velocity curve and the blood flow velocity threshold; The segmented blood flow velocity area image determination unit is used to identify the red blood cell aggregates on the blood flow velocity area image based on the region growing algorithm, and set the identified red blood cell aggregates to white, and set the area outside the red blood cell aggregates to black, so as to obtain the segmented blood flow velocity area image; the blood flow velocity area image includes a blood high flow velocity area image and a blood low flow velocity area image; A feature extraction unit is used to extract features from the segmented blood flow velocity area image to obtain image features; the image features include the number of white independent connected domains, the area of ​​the white independent connected domains, the perimeter of the white independent connected domains, the area of ​​the circumscribed ellipse of the white independent connected domains, the area of ​​the black area, and the total area of ​​the segmented blood flow velocity area image; The red blood cell aggregation value determination unit is used to calculate the red blood cell aggregation value in the blood flow rate area image based on the image features and using the red blood cell aggregation calculation formula.