A medical image segmentation method for intracranial arteriosclerotic plaques in patients with atrial fibrillation

By analyzing edge morphology and grayscale changes in intracranial MRA images, the probability of plaque existence is calculated and the grayscale value is adjusted to generate enhanced images, which solves the problem of inaccurate segmentation caused by low contrast between plaque and blood vessels, and achieves a more accurate Plaque segmentation and diagnostic support.

CN118379313BActive Publication Date: 2024-11-12BEIJING HOSPITAL +2
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410694466.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-05-31
Publication Date
2024-11-12
Estimated Expiration
2044-05-31

AI Technical Summary

Technical Problem

In the prior art, the medical image segmentation results of intracranial atherosclerotic plaques have poor accuracy, mainly due to the low contrast between the plaque and blood vessels, resulting in inaccurate misjudgment and segmentation.

Method used

By obtaining the grayscale image of intracranial MRA, the edge morphological characteristics, grayscale fluctuation characteristics and local grayscale variation in the blood vessel area are used to calculate the probability of plaques in each pixel point, and the grayscale value is adjusted to generate intracranial MRA Grayscale enhanced image, Otsu method threshold segmentation algorithm is used for plaque segmentation.

Benefits of technology

It improves the contrast between the plaque and the vascular junction edge, enhances the accuracy of segmentation results, reduces misjudgment, provides more accurate plaque location and morphological information, and supports more effective diagnosis and treatment.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118379313B_ABST
    Figure CN118379313B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of plaque segmentation technology, and specifically to a medical image segmentation method for intracranial arteriosclerotic plaques in patients with atrial fibrillation. The method obtains multiple vascular segmentation areas; obtains the probability of plaque existence in each vascular segmentation area according to the grayscale fluctuation characteristics of the pixel points in each vascular segmentation area, the morphological change characteristics of the corresponding vascular segmentation area, and the neighborhood grayscale distribution; obtains the possibility that the pixel point belongs to the boundary edge between the plaque and the blood vessel according to the gradient characteristics of each pixel point in each vascular segmentation area, the local grayscale change degree, and the corresponding plaque existence probability; adjusts the grayscale value of the pixel point, obtains the enhanced grayscale value of each pixel point, and obtains the intracranial MRA grayscale enhanced image; and segments the plaque. The present invention improves the accuracy of plaque segmentation by adaptively enhancing the grayscale value of the pixel points at the boundary edge between the plaque and the blood vessel.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of plaque segmentation, and in particular to a medical image segmentation method for intracranial arteriosclerotic plaques in patients with atrial fibrillation. Background Art

[0002] Atherosclerotic plaques are one of the important manifestations of intracranial vascular diseases. Image processing technology can be used to accurately segment intracranial arteriosclerotic plaques, so that the size and distribution of plaques can be more clearly understood and the degree of their impact on blood vessels can be evaluated. At the same time, for the long-term observation of intracranial arteriosclerosis, plaque segmentation technology based on image processing can provide more accurate and reliable data support.

[0003] In the prior art, threshold segmentation is used to achieve a simple division of the grayscale of vascular images. However, since medical images of intracranial arteriosclerotic plaques usually have complex grayscale distribution and contrast changes, the presence of plaques leads to a decrease in blood flow in regional blood vessels, and the contrast between plaques and blood vessels in the image is low, the accuracy of plaque segmentation results is poor. Summary of the invention

[0004] In order to solve the technical problem that the contrast between plaque and blood vessel is low and the accuracy of plaque segmentation result is poor, the purpose of the present invention is to provide a medical image segmentation method for intracranial arteriosclerotic plaque in patients with atrial fibrillation. The technical scheme adopted is as follows:

[0005] The present invention proposes a medical image segmentation method for intracranial arteriosclerotic plaques in patients with atrial fibrillation, the method comprising:

[0006] Acquire an intracranial MRA grayscale image of a patient with atrial fibrillation; the intracranial MRA grayscale image includes a vascular area;

[0007] According to the edge morphological features of the neighborhood of each edge pixel point in the vascular region, multiple vascular segmentation regions are obtained; according to the grayscale fluctuation features of the pixels in each vascular segmentation region, the morphological change features of the corresponding vascular segmentation region and the neighborhood grayscale distribution, the probability of plaque existence in each vascular segmentation region is obtained;

[0008] According to the gradient characteristics of each pixel point in each blood vessel segmentation area, the local grayscale change degree and the corresponding plaque existence probability, the possibility of each pixel point belonging to the boundary edge between the plaque and the blood vessel is obtained; according to the possibility of each pixel point, the grayscale value of each pixel point is adjusted to obtain the enhanced grayscale value of each pixel point, and the intracranial MRA grayscale enhanced image is obtained;

[0009] The plaque is segmented according to the intracranial MRA grayscale enhanced image.

[0010] Furthermore, the method for obtaining the blood vessel segmentation region includes:

[0011] Traverse the neighborhood range of each edge pixel point in the blood vessel area and construct the minimum circumscribed rectangle of the neighborhood range of each pixel point; for any edge pixel point in the neighborhood range, obtain the direction of the line connecting the edge pixel point to two adjacent edge pixels, and calculate the angle between each connecting line direction and the long direction of the minimum circumscribed rectangle as the direction angle;

[0012] Calculate the difference between the angles of the two connecting line directions corresponding to the directions as the angle difference;

[0013] If the angle difference of the edge pixel points is greater than the preset angle, and the angle difference of the edge pixel points is greater than the angle difference of adjacent edge pixel points, the edge pixel points are marked as segmentation pixels;

[0014] The neighborhood range of edge pixel points is segmented along the normal direction of the segmentation pixel points in the extending direction of the blood vessel to obtain multiple blood vessel segmentation areas.

[0015] Furthermore, the method for obtaining the plaque existence probability includes:

[0016] Calculate the variance of the grayscale values ​​of all pixels in each blood vessel segmentation area as the grayscale fluctuation feature;

[0017] Calculate the relative distance between the edge pixel points in the extension direction of each blood vessel segment area and other edge pixel points in the corresponding normal direction, and calculate the difference between the maximum relative distance and the minimum relative distance as the morphological change feature;

[0018] According to the neighborhood grayscale distribution of the blood vessel segmentation region, the grayscale difference feature between the blood vessel segmentation region and the adjacent blood vessel segmentation region in the corresponding extension direction is obtained;

[0019] The product of the grayscale fluctuation feature, the morphological change feature and the grayscale difference feature is calculated as the probability of plaque existence in each blood vessel segment area.

[0020] Furthermore, the method for acquiring the grayscale difference feature includes:

[0021] Calculate the mean grayscale value of all pixels in each vascular segmentation area as the first mean; calculate the mean grayscale value of all pixels in each adjacent vascular segmentation area in the extension direction of each vascular segmentation area as the second mean; calculate the difference between the second mean and the first mean as the first difference; calculate the mean of the first difference between each vascular segmentation area and all adjacent vascular segmentation areas in the corresponding extension direction as the grayscale difference feature.

[0022] Furthermore, the method for obtaining the local grayscale variation degree includes:

[0023] For each pixel point in the blood vessel segmentation area, a preset number of pairs of adjacent pixel points are obtained on both sides of the normal line in the gradient direction;

[0024] The difference in grayscale values ​​between each pair of adjacent pixel points is calculated as the grayscale difference; and the average of the grayscale differences between all pairs of adjacent pixel points is calculated as the local grayscale variation.

[0025] Furthermore, the method for obtaining the possibility includes:

[0026] The possibility of each pixel belonging to the boundary edge between the plaque and the blood vessel is obtained according to the possibility acquisition formula. The possibility acquisition formula is:

[0027] ;in, Indicates the first The probability that a pixel belongs to the boundary between plaque and blood vessel; Indicates The pixel is located at The probability of plaque presence in each vascular segment area; Indicates The pixel is located at Minimum value of pixel gradient in each blood vessel segmentation area; Indicates The pixel is located at The maximum value of pixel gradient in each blood vessel segment area; Indicates The gradient value of each pixel; Indicates The local grayscale variation of each pixel; Indicates taking the absolute value.

[0028] Furthermore, the method for obtaining the enhanced grayscale value includes:

[0029] Normalize the possibility of each pixel belonging to the boundary edge of the plaque and the blood vessel, and calculate the product of the normalized result and the gray value of the corresponding pixel as the gray weighted value;

[0030] The sum of the grayscale weighted value and the grayscale value is calculated to obtain the enhanced grayscale value of each pixel.

[0031] Furthermore, segmenting the plaque according to the intracranial MRA grayscale enhanced image includes:

[0032] The plaques were obtained by using Otsu threshold segmentation algorithm on the gray-scale enhanced images of intracranial MRA.

[0033] Furthermore, the method for acquiring the blood vessel region includes:

[0034] The intracranial MRA grayscale image was segmented by local adaptive threshold to obtain a binary image; the intracranial MRA grayscale image was detected by Canny edge detection to obtain an edge image; the binary image and the edge image were XOR-fused to obtain the vascular area.

[0035] Furthermore, the relative distance is obtained by calculating the Euclidean distance.

[0036] The present invention has the following beneficial effects:

[0037] The present invention obtains multiple vascular segmentation regions according to the edge morphological features within the neighborhood range of each edge pixel point in the vascular region, can more accurately capture local changes, and obtain more accurate and practical vascular segmentation regions; according to the grayscale fluctuation characteristics of the pixel points in each vascular segmentation region, the morphological change characteristics of the corresponding vascular segmentation region and the neighborhood grayscale distribution, the probability of plaque existence in each vascular segmentation region is obtained, and it is more accurately judged whether there is a plaque in the vascular segmentation region, and it is avoided that a non-plaque region is misjudged as a plaque region; according to the gradient characteristics of each pixel point in each vascular segmentation region, the local grayscale change degree and the corresponding plaque existence probability, the possibility of each pixel point belonging to the boundary edge between the plaque and the blood vessel is obtained, the possibility of the pixel point at the boundary edge between the plaque and the blood vessel is evaluated, and the position of the boundary edge is accurately depicted; the grayscale value of each pixel point is adjusted according to the possibility of each pixel point, the enhanced grayscale value of each pixel point is obtained, and the intracranial MRA grayscale enhanced image is obtained, the grayscale value of the pixel point is adjusted more specifically, the contrast between the blood vessel and the plaque is more obvious, and it is helpful to observe and identify the vascular structure more clearly; the plaque is segmented. The present invention improves the accuracy of plaque segmentation by adaptively enhancing the grayscale value of pixel points at the boundary edge between the plaque and the blood vessel. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0039] Figure 1 A flowchart of a medical image segmentation method for intracranial arteriosclerotic plaque in patients with atrial fibrillation provided by one embodiment of the present invention;

[0040] Figure 2 A schematic diagram of an intracranial MRA image containing arteriosclerotic plaque provided by an embodiment of the present invention;

[0041] Figure 3A schematic diagram of a binary image of an intracranial MRA grayscale image provided by an embodiment of the present invention;

[0042] Figure 4 A schematic diagram of an edge image of an intracranial MRA grayscale image provided by an embodiment of the present invention;

[0043] Figure 5 A schematic diagram of determining segmentation pixels when there is blood vessel overlap or plaque in a blood vessel region where an edge pixel is located, provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0044] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation method, structure, features and effects of a medical image segmentation method for intracranial arteriosclerotic plaques in patients with atrial fibrillation proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0045] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0046] The specific scheme of the medical image segmentation method for intracranial arteriosclerotic plaque in patients with atrial fibrillation provided by the present invention is described in detail below with reference to the accompanying drawings.

[0047] See also Figure 1 , which shows a flow chart of a medical image segmentation method for intracranial arteriosclerotic plaques in patients with atrial fibrillation provided by an embodiment of the present invention, and the specific method includes:

[0048] Step S1: Acquire an intracranial MRA grayscale image of a patient with atrial fibrillation; the intracranial MRA grayscale image includes a vascular region.

[0049] In the embodiment of the present invention, in order to accurately identify the location, size and morphology of plaques, thereby improving the diagnostic accuracy of intracranial arteriosclerotic plaques and guiding the selection of treatment plans, it is necessary to accurately segment the intracranial arteriosclerotic plaques of patients with atrial fibrillation; the diagnosis of intracranial arteriosclerosis plaques in patients with atrial fibrillation usually uses magnetic resonance angiography (MRA) technology to obtain detailed images of the patient's intracranial blood vessels; first, a magnetic resonance imaging (MRI) machine is used for patients with atrial fibrillation to detect the movement of water molecules in the body based on magnetic fields and radio frequency pulses, and the blood vessel structure and blood flow state are reconstructed by measuring the signals generated by blood flow to obtain intracranial magnetic resonance angiography (MRA) images; Figure 2, shows a schematic diagram of an intracranial MRA image containing an arteriosclerotic plaque, with the arrow pointing to the intracranial arteriosclerotic plaque.

[0050] In one embodiment of the present invention, in order to facilitate the subsequent image processing process, the acquired intracranial MRA image is preprocessed to enhance the image quality and obtain an intracranial MRA grayscale image of a patient with atrial fibrillation. It should be noted that the image preprocessing operation is a technical means well known to those skilled in the art, and can be specifically set according to the specific implementation scenario. In one embodiment of the present invention, a grayscale algorithm and Gaussian filtering are used for preprocessing to obtain an intracranial MRA grayscale image, which can highlight the contour and details of the image, making the image clearer and easier to perform operations such as feature extraction and image recognition. The specific grayscale algorithm and Gaussian filtering are technical means well known to those skilled in the art, and will not be described in detail here.

[0051] Since there is a large amount of information in the intracranial MRA grayscale image, which affects the subsequent processing, in order to separate the region of interest as much as possible and simplify the image information, the vascular area is analyzed.

[0052] It should be noted that, in one embodiment of the present invention, the method for acquiring the blood vessel region includes:

[0053] Perform local adaptive threshold segmentation on the intracranial MRA grayscale image to obtain a binary image, such as Figure 3 , shows a schematic diagram of a binary image of an intracranial MRA grayscale image; Canny edge detection is performed on the intracranial MRA grayscale image to obtain an edge image, such as Figure 4 , showing a schematic diagram of an edge image of an intracranial MRA grayscale image; the binary image and the edge image are XOR-fused to obtain the vascular area.

[0054] By selecting a suitable threshold, blood vessels can be distinguished from surrounding tissues, making the blood vessel structure more prominent in the image; in MRA grayscale images, the edges of blood vessels usually have obvious brightness changes. Through edge detection algorithms, the edges of blood vessels can be further highlighted, making the contours of blood vessels clearer; by combining threshold segmentation and edge detection methods, blood vessel areas can be extracted more accurately, reducing false detection and missed detection. Specific local adaptive threshold segmentation, Canny edge detection, and XOR fusion are technical means well known to those skilled in the art, and will not be described in detail here.

[0055] Step S2: Based on the edge morphological features within the neighborhood of each edge pixel point in the vascular region, multiple vascular segmentation regions are obtained; based on the grayscale fluctuation characteristics of the pixels in each vascular segmentation region, the morphological change characteristics of the corresponding vascular segmentation region, and the neighborhood grayscale distribution, the probability of plaque existence in each vascular segmentation region is obtained.

[0056] Blood vessels present complex and changeable shapes. Blood vessels in different regions may have different degrees of curvature, bifurcation, and width changes. By analyzing the edge morphological features within the neighborhood of each edge pixel point, these local changes can be captured more accurately, and adaptive adjustments can be made according to the actual changes in the blood vessel area to obtain more accurate and practical blood vessel segmentation areas. Therefore, multiple blood vessel segmentation areas are obtained based on the edge morphological features within the neighborhood of each edge pixel point in the blood vessel area.

[0057] Preferably, in one embodiment of the present invention, the method for acquiring the blood vessel segmentation region includes:

[0058] Traverse the neighborhood range of each edge pixel point in the blood vessel area and construct the minimum circumscribed rectangle of the neighborhood range of each pixel point; for any edge pixel point in the neighborhood range, obtain the direction of the line connecting the edge pixel point to two adjacent edge pixels, and calculate the angle between each connecting line direction and the long direction of the minimum circumscribed rectangle as the direction angle;

[0059] Calculate the difference between the angles of the two connecting line directions corresponding to the directions as the angle difference;

[0060] If the angle difference of the edge pixel points is greater than the preset angle, and the angle difference of the edge pixel points is greater than the angle difference of the adjacent edge pixel points, the edge pixel points are marked as segmentation pixels; Figure 5 , shows a schematic diagram of determining segmentation pixels when there is vascular overlap or plaque in the vascular region where the edge pixel is located. The left figure shows that there is vascular overlap in the vascular region where the edge pixel is located, and the right figure shows that there is plaque in the vascular region where the edge pixel is located. The direction of the arrows in the figure is the direction of the connecting line, and the intersection point between the arrows is the segmentation pixel point.

[0061] The neighborhood range of edge pixel points is segmented along the normal direction of the segmentation pixel points in the extending direction of the blood vessel to obtain multiple blood vessel segmentation areas.

[0062] It should be noted that, in one embodiment of the present invention, due to the complex structure of intracranial blood vessels, localized analysis of the neighborhood range of each edge pixel point is helpful to capture subtle changes in blood vessel morphology and structure. The method for obtaining the neighborhood range of the edge pixel point is as follows: for the edge pixel point gradient direction, the preset times of the number of pixels in the blood vessel area are used as the length of the neighborhood range, and the range in the blood vessel area constitutes the neighborhood range of the edge pixel point, wherein the preset times is 3 times; the preset angle is 5°; in other embodiments of the present invention, the size of the neighborhood range, the preset times and the preset angle can be set according to the specific situation, and will not be limited or elaborated here.

[0063] The morphological change feature describes the shape and structural changes of the vascular segmentation area. When plaques exist, the morphology of the blood vessels in the image is concave, causing the vascular channel to narrow, and the greater the morphological change; the grayscale fluctuation feature reflects the change in the grayscale value of the pixel points in the vascular segmentation area. Plaques reduce the blood flow of the blood vessels and the signal emitted becomes smaller, resulting in a smaller grayscale value in the existing area and changes, and the grayscale fluctuation feature is larger; under normal circumstances, the grayscale values ​​between vascular segmentation areas are close and the difference is small, but the presence of plaques will cause the grayscale value of the area to become smaller, the grayscale difference becomes larger, and the probability of plaque existence is greater. Therefore, according to the grayscale fluctuation characteristics of the pixels in each vascular segmentation area, the morphological change characteristics of the corresponding vascular segmentation area, and the grayscale distribution of the neighborhood, the probability of plaque existence in each vascular segmentation area is obtained.

[0064] Preferably, in one embodiment of the present invention, the method for obtaining the plaque existence probability includes:

[0065] Calculate the variance of the grayscale values ​​of all pixels in each blood vessel segmentation area as the grayscale fluctuation feature;

[0066] Calculate the relative distance between the edge pixel points in the extension direction of each blood vessel segment area and other edge pixel points in the corresponding normal direction, and calculate the difference between the maximum relative distance and the minimum relative distance as the morphological change feature;

[0067] According to the neighborhood grayscale distribution of the blood vessel segmentation region, the grayscale difference feature between the blood vessel segmentation region and the adjacent blood vessel segmentation region in the corresponding extension direction is obtained;

[0068] The product of the grayscale fluctuation feature, the morphological change feature, and the grayscale difference feature is calculated as the probability of plaque existence in each vascular segment area.

[0069] Preferably, in one embodiment of the present invention, the method for acquiring grayscale difference features includes:

[0070] Calculate the mean grayscale value of all pixels in each vascular segmentation area as the first mean; calculate the mean grayscale value of all pixels in each adjacent vascular segmentation area in the extension direction of each vascular segmentation area as the second mean; calculate the difference between the second mean and the first mean as the first difference; calculate the mean of the first difference between each vascular segmentation area and all adjacent vascular segmentation areas in the corresponding extension direction as the grayscale difference feature.

[0071] In one embodiment of the present invention, the formula for the probability of plaque existence is expressed as:

[0072] ;

[0073] in, Indicates The probability of plaque presence in each vascular segment area; Indicates The number of pixels in each blood vessel segmentation area; Indicates The sequence number of the pixel points in the blood vessel segmentation area; Indicates In the vessel segment area The gray value of each pixel; Indicates The mean gray value of all pixels in the blood vessel segmentation area; Indicates The difference between the maximum relative distance and the minimum relative distance in each blood vessel segment area; Indicates The number of adjacent vascular segmentation regions in the extension direction of a vascular segmentation region; Indicates The serial numbers of the adjacent blood vessel segmentation regions in the extension direction of the blood vessel segmentation region; Indicates The first segment of the blood vessel The gray value average of all pixels in the adjacent blood vessel segmentation area.

[0074] In the formula for the probability of plaque existence, The formula for finding the variance of the grayscale values ​​of all pixels in each blood vessel segment area is expressed as The grayscale fluctuation characteristics of the pixels in the blood vessel segmentation area are as follows: the larger the grayscale fluctuation characteristics, the greater the grayscale change of the pixels in the blood vessel segmentation area, and the more likely it is that there is a plaque; the smaller the grayscale fluctuation characteristics, the more uniform the grayscale change of the pixels in the blood vessel segmentation area, and the more normal the blood vessel area. Indicates the morphological change characteristics. The larger the morphological change characteristics, the greater the change in blood vessel width, and the more likely it is that the plaque exists in the area; It is the grayscale difference feature between each vascular segmentation area and all adjacent vascular segmentation areas in the corresponding extension direction. The larger the grayscale difference feature, the smaller the grayscale mean of the vascular segmentation area. The greater the difference in grayscale mean between each vascular segmentation area and all adjacent vascular segmentation areas in the corresponding extension direction, the more likely it is a plaque existence area. On the contrary, the smaller the grayscale difference feature, the smaller the difference in grayscale mean between each vascular segmentation area and all adjacent vascular segmentation areas in the corresponding extension direction, the greater the possibility of being a normal area. Therefore, the larger the grayscale fluctuation feature, the larger the morphological change feature, the larger the grayscale difference feature, the greater the possibility of belonging to a plaque existence area, and the greater the probability of plaque existence in the vascular segmentation area.

[0075] It should be noted that, in one embodiment of the present invention, the method for obtaining the relative distance is to calculate the Euclidean distance, and the interval for obtaining the relative distance is 1 pixel; in other embodiments of the present invention, other basic mathematical operations can also be used to obtain the relative distance. The specific method is a technical means well known to those skilled in the art and will not be elaborated here.

[0076] Step S3: According to the gradient characteristics of each pixel point in each vascular segmentation area, the local grayscale change and the corresponding probability of plaque existence, the possibility of each pixel point belonging to the boundary edge between the plaque and the blood vessel is obtained; according to the possibility of each pixel point, the grayscale value of each pixel point is adjusted to obtain the enhanced grayscale value of each pixel point, and the intracranial MRA grayscale enhanced image is obtained.

[0077] Gradient feature is an important indicator to measure the local structural change of the image. At the boundary edge between plaque and blood vessel, due to the large structural difference, the gradient change of pixel points is usually more significant, and the larger the gradient feature; at the boundary edge between blood vessel and plaque, due to the difference in tissue type and density, the grayscale values ​​on both sides of the boundary edge often change significantly. Therefore, by calculating the local grayscale change of the pixel point, it can be further confirmed whether the pixel point is located on the boundary edge. The larger the local grayscale change, the more likely the pixel point is at the junction of different tissue types; the greater the probability of plaque existence in the vascular segmentation area, the greater the credibility of the pixel point belonging to the boundary edge between the plaque and the blood vessel; therefore, according to the gradient feature, local grayscale change and corresponding plaque existence probability of each pixel point in each vascular segmentation area, the possibility of each pixel point belonging to the boundary edge between the plaque and the blood vessel is obtained.

[0078] Preferably, in one embodiment of the present invention, the method for obtaining the local grayscale variation degree includes:

[0079] For each pixel point in the blood vessel segmentation area, a preset number of pairs of adjacent pixel points are obtained on both sides of the normal line in the gradient direction; the difference in grayscale values ​​between each pair of adjacent pixel points is calculated as the grayscale difference; the mean of the grayscale differences between all pairs of adjacent pixel points is calculated as the local grayscale variation.

[0080] In one embodiment of the present invention, the formula for the local grayscale variation is expressed as:

[0081] ;

[0082] in, Indicates The local grayscale variation of each pixel; Indicates The serial numbers of the adjacent pixels on both sides of the normal line of a pixel in the gradient direction; Indicates The normal of a pixel in the gradient direction Side The gray value of each pixel; Indicates The normal of a pixel in the gradient direction Side The gray value of each pixel; Indicates The number of adjacent pixels selected on both sides of the normal line of the gradient direction of a pixel.

[0083] In the formula of local grayscale variation, the greater the difference in grayscale values ​​between adjacent pixels on both sides of the normal line, the more the two sides are not in the same area, and the more likely they are to belong to the boundary between the plaque and the blood vessel; the smaller the difference in grayscale values ​​between adjacent pixels on both sides of the normal line, the more likely they are to be the same type of area, and the smaller the local variation.

[0084] It should be noted that, in one embodiment of the present invention, the preset number is 3, that is, 3 pairs of adjacent pixel points are selected on both sides of the normal line; in other embodiments of the present invention, the size of the preset number can be set according to the specific situation, and is not limited or elaborated here.

[0085] Preferably, in one embodiment of the present invention, the method for obtaining the possibility includes:

[0086] The possibility of each pixel belonging to the boundary edge between the plaque and the blood vessel is obtained according to the possibility acquisition formula. The possibility acquisition formula is:

[0087] ;

[0088] in, Indicates the first The probability that a pixel belongs to the boundary between plaque and blood vessel; Indicates The pixel is located at The probability of plaque presence in each vascular segment area; Indicates The pixel is located at Minimum value of pixel gradient in each blood vessel segmentation area; Indicates The pixel is located at The maximum value of pixel gradient in each blood vessel segment area; Indicates The gradient value of each pixel; Indicates The local grayscale variation of each pixel; Indicates taking the absolute value.

[0089] In the probability acquisition formula, Indicates The pixel is located at The minimum gradient of the blood vessel segment area is The difference in gradient values ​​of pixels is greater when the difference is larger. The larger the gradient value of a pixel point is, the greater the possibility that the pixel point belongs to the boundary between the plaque and the blood vessel; Indicates The pixel is located at The maximum gradient of the blood vessel segment area is The gradient value difference of the pixel points is smaller, the The closer the gradient value of a pixel is to the maximum value, the greater the possibility that the pixel belongs to the boundary between the plaque and the blood vessel; It is to prevent the When the gradient value of a pixel is equal to the maximum value, the denominator of the formula is 0, which makes the formula meaningless. Represents the ratio of two differences. The larger the ratio, the The larger the gradient value of the pixel point, the greater the possibility that the pixel point belongs to the boundary between the plaque and the blood vessel; The pixel is located at The greater the probability of plaque existence in the first blood vessel segmentation area, the greater the confidence in the subsequent judgment of whether the pixel point belongs to the boundary edge between the plaque and the blood vessel; The greater the local grayscale variation of a pixel, the greater the grayscale difference of the pixel on both sides of the normal line in the gradient direction, and the more likely the pixel belongs to the boundary edge between the plaque and the blood vessel.

[0090] It should be noted that in other embodiments of the present invention, in order to more accurately locate and analyze the boundary edge between the blood vessel and the plaque, the processing range is limited to the minimum area containing the blood vessel segmentation area, thereby reducing misjudgment due to morphological changes or inconsistent directions and improving the accuracy of boundary edge detection. Implementers may also construct a minimum circumscribed rectangle for each blood vessel segmentation area and analyze the possibility that each pixel point within the minimum circumscribed rectangle belongs to the boundary edge between the plaque and the blood vessel.

[0091] Grayscale values ​​represent the original brightness and color information of pixels, and reflect the structure and characteristics of different regions in the image. The greater the possibility that a pixel belongs to the border between plaques and blood vessels, the more likely it is that there will be changes in the regional structure. The more important the analysis of different regions is, the more it is necessary to enhance the grayscale value. By enhancing the grayscale value, the contrast between blood vessels and plaques can be made more obvious, which helps to observe and identify the vascular structure more clearly. Therefore, the grayscale value of each pixel is adjusted according to the possibility of each pixel, the enhanced grayscale value of each pixel is obtained, and the intracranial MRA grayscale enhanced image is obtained.

[0092] Preferably, in one embodiment of the present invention, the method for acquiring the enhanced gray value includes:

[0093] Normalize the probability of each pixel belonging to the boundary edge of the plaque and the blood vessel, calculate the product of the normalized result and the gray value of the corresponding pixel, and use it as the gray weighted value; calculate the sum of the gray weighted value and the gray value to obtain the enhanced gray value of each pixel. In one embodiment of the present invention, the formula for the enhanced gray value is expressed as:

[0094] ;

[0095] in, Indicates The enhanced gray value of each pixel; Indicates The gray value of each pixel; Indicates The probability that a pixel belongs to the boundary between plaque and blood vessel; Indicates The result of normalizing the probability that a pixel belongs to the boundary between plaque and blood vessel.

[0096] In the formula for enhancing grayscale value, Indicates The grayscale of each pixel is enhanced according to the possibility of belonging to the boundary between plaque and blood vessel. Represents the grayscale weighted value. The greater the possibility that each pixel belongs to the boundary edge of the plaque and the blood vessel, the larger the grayscale value and the grayscale weighted value. The larger the grayscale value of the pixel after enhancement, the more prominent the contrast between the plaque and the blood vessel.

[0097] It should be noted that, in one embodiment of the present invention, the possibility of a pixel point belonging to the boundary edge of a plaque and a blood vessel can be normalized by using basic mathematical operations such as maximum and minimum normalization and standard deviation normalization. The specific method is a technical means well known to those skilled in the art and will not be elaborated here.

[0098] Step S4: Segment the plaque according to the intracranial MRA grayscale enhanced image.

[0099] After obtaining the enhanced grayscale value of the pixel point, the enhancement operation of the intracranial MRA grayscale image is completed, which effectively improves the accuracy of the subsequent image processing results, highlights the contrast of the boundary edge between the enlarged blood vessels and plaques, improves the integrity of plaque segmentation, and segments the plaque according to the intracranial MRA grayscale enhanced image.

[0100] Preferably, in one embodiment of the present invention, segmenting the plaque according to the intracranial MRA grayscale enhanced image includes:

[0101] The plaques were obtained by using Otsu threshold segmentation algorithm on the gray-scale enhanced images of intracranial MRA.

[0102] By accurately segmenting the plaque, it helps to provide doctors with clearer information about the size, location, shape and relationship of the plaque with surrounding blood vessels, which is conducive to more accurate diagnosis and treatment plan formulation. It should be noted that the specific Otsu threshold segmentation algorithm is a technical means well known to those skilled in the art, and is not limited or elaborated here.

[0103] In summary, the present invention obtains multiple vascular segmentation regions; obtains the probability of plaque existence in each vascular segmentation region according to the grayscale fluctuation characteristics of the pixel points in each vascular segmentation region, the morphological change characteristics of the corresponding vascular segmentation region, and the grayscale distribution of the neighborhood; obtains the possibility that the pixel point belongs to the boundary edge between the plaque and the blood vessel according to the gradient characteristics of each pixel point in each vascular segmentation region, the local grayscale change degree, and the corresponding plaque existence probability; adjusts the grayscale value of each pixel point, obtains the enhanced grayscale value of each pixel point, and obtains the intracranial MRA grayscale enhanced image; and segments the plaque. The present invention improves the accuracy of plaque segmentation by adaptively enhancing the grayscale value of the pixel points at the boundary edge between the plaque and the blood vessel.

[0104] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0105] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A medical image segmentation method for intracranial arteriosclerotic plaques in patients with atrial fibrillation, characterized in that: The method comprises: Acquire an intracranial MRA grayscale image of a patient with atrial fibrillation; the intracranial MRA grayscale image includes a vascular area; According to the edge morphological features of the neighborhood of each edge pixel point in the vascular region, multiple vascular segmentation regions are obtained; according to the grayscale fluctuation features of the pixels in each vascular segmentation region, the morphological change features of the corresponding vascular segmentation region and the neighborhood grayscale distribution, the probability of plaque existence in each vascular segmentation region is obtained; According to the gradient characteristics of each pixel point in each blood vessel segmentation area, the local grayscale change degree and the corresponding plaque existence probability, the possibility of each pixel point belonging to the boundary edge between the plaque and the blood vessel is obtained; according to the possibility of each pixel point, the grayscale value of each pixel point is adjusted to obtain the enhanced grayscale value of each pixel point, and the intracranial MRA grayscale enhanced image is obtained; Segmenting the plaque according to the intracranial MRA grayscale enhanced image; The method for obtaining the plaque existence probability includes: Calculate the variance of the grayscale values ​​of all pixels in each blood vessel segmentation area as the grayscale fluctuation feature; Calculate the relative distance between the edge pixel points in the extension direction of each blood vessel segment area and other edge pixel points in the corresponding normal direction, and calculate the difference between the maximum relative distance and the minimum relative distance as the morphological change feature; According to the neighborhood grayscale distribution of the blood vessel segmentation region, the grayscale difference feature between the blood vessel segmentation region and the adjacent blood vessel segmentation region in the corresponding extension direction is obtained; Calculating the product of the grayscale fluctuation feature, the morphological change feature and the grayscale difference feature as the probability of plaque existence in each blood vessel segment area; The method for obtaining the grayscale difference feature includes: Calculate the mean grayscale value of all pixels in each blood vessel segmentation area as the first mean; calculate the mean grayscale value of all pixels in each adjacent blood vessel segmentation area in the extension direction of each blood vessel segmentation area as the second mean; calculate the difference between the second mean and the first mean as the first difference; calculate the mean of the first difference between each blood vessel segmentation area and all adjacent blood vessel segmentation areas in the corresponding extension direction as the grayscale difference feature; The method for obtaining the possibility includes: The possibility of each pixel belonging to the boundary edge between the plaque and the blood vessel is obtained according to the possibility acquisition formula. The possibility acquisition formula is: ;in, Indicates the first The probability that a pixel belongs to the boundary between plaque and blood vessel; Indicates The pixel is located at The probability of plaque presence in each vascular segment area; Indicates The pixel is located at Minimum value of pixel gradient in each blood vessel segmentation area; Indicates The pixel is located at The maximum value of pixel gradient in each blood vessel segment area; Indicates The gradient value of each pixel; Indicates The local grayscale variation of each pixel; Indicates taking the absolute value; The method for obtaining the enhanced grayscale value comprises: Normalize the possibility of each pixel belonging to the boundary edge of the plaque and the blood vessel, and calculate the product of the normalized result and the gray value of the corresponding pixel as the gray weighted value; The sum of the grayscale weighted value and the grayscale value is calculated to obtain the enhanced grayscale value of each pixel.

2. The method for segmenting intracranial arteriosclerotic plaques in patients with atrial fibrillation according to claim 1, characterized in that: The method for obtaining the blood vessel segmentation region comprises: Traverse the neighborhood range of each edge pixel point in the blood vessel area and construct the minimum circumscribed rectangle of the neighborhood range of each pixel point; for any edge pixel point in the neighborhood range, obtain the direction of the line connecting the edge pixel point to two adjacent edge pixels, and calculate the angle between each connecting line direction and the long direction of the minimum circumscribed rectangle as the direction angle; Calculate the difference between the angles of the two connecting line directions corresponding to the directions as the angle difference; If the angle difference of the edge pixel points is greater than the preset angle, and the angle difference of the edge pixel points is greater than the angle difference of adjacent edge pixel points, the edge pixel points are marked as segmentation pixels; The neighborhood range of edge pixel points is segmented along the normal direction of the segmentation pixel points in the extending direction of the blood vessel to obtain multiple blood vessel segmentation areas.

3. The medical image segmentation method for intracranial arteriosclerotic plaque in patients with atrial fibrillation according to claim 1, characterized in that: The method for obtaining the local grayscale variation degree includes: For each pixel point in the blood vessel segmentation area, a preset number of pairs of adjacent pixel points are obtained on both sides of the normal line in the gradient direction; The difference in grayscale values ​​between each pair of adjacent pixel points is calculated as the grayscale difference; and the average of the grayscale differences between all pairs of adjacent pixel points is calculated as the local grayscale variation.

4. The method for segmenting intracranial arteriosclerotic plaques in patients with atrial fibrillation according to claim 1, characterized in that: The segmenting of the plaque according to the intracranial MRA grayscale enhanced image comprises: The plaques were obtained by using Otsu threshold segmentation algorithm on the gray-scale enhanced images of intracranial MRA.

5. The method for segmenting medical images of intracranial arteriosclerotic plaques in patients with atrial fibrillation according to claim 1, characterized in that: The method for acquiring the blood vessel region comprises: The intracranial MRA grayscale image was segmented by local adaptive threshold to obtain a binary image; the intracranial MRA grayscale image was detected by Canny edge detection to obtain an edge image; the binary image and the edge image were XOR-fused to obtain the vascular area.

6. The method for segmenting intracranial arteriosclerotic plaques in patients with atrial fibrillation according to claim 1, characterized in that: The method for obtaining the relative distance is to calculate the Euclidean distance.