A method and system for locating cerebrovascular lesions based on image recognition

Abstract

The present invention relates to the field of cerebrovascular image recognition, and specifically to a method and system for locating cerebrovascular lesions based on image recognition. The present invention obtains cerebrovascular edge pixel points, establishes a preset sliding window with the edge pixel points as the starting point, and traverses the cerebrovascular area; obtains the degree of cerebrovascular expansion according to the pixel point distribution of the cerebrovascular area; obtains the expansion coefficient in combination with the pixel point distribution of the cerebrovascular area in the sliding window area; obtains the abnormal coefficient according to the difference in expansion coefficient between the sliding window area and other sliding window areas; screens out abnormal cerebrovascular areas according to the abnormal coefficient; and then locates cerebrovascular lesions. The present invention avoids the errors that may be caused by the ambiguity and complexity of the cerebrovascular image itself, thereby enabling relevant personnel to accurately locate the position of cerebrovascular lesions.

Description

Technical Field

[0001] The present invention relates to the field of cerebrovascular image recognition, and in particular to a cerebrovascular lesion positioning method and system based on image recognition. Background Art

[0002] Digital subtraction angiography (DSA) technology is an important method for evaluating cerebral aneurysms and has been widely used in actual clinical practice. As the gold standard for the diagnosis and treatment effect evaluation of cerebrovascular diseases, DSA technology is known for its high sensitivity, high specificity and high accuracy. This technology has excellent spatial resolution and dynamically displays the actual situation of whole-brain blood flow, allowing doctors to deeply understand the severity and scope of target vascular lesions and provide reliable and objective basis for subsequent treatment. Combined with image processing technology, DSA technology can optimize the efficiency of locating cerebrovascular lesions. Image correlation technology can improve the clarity and contrast of images and enhance the visualization of lesions, thereby helping doctors identify lesions more accurately. In addition, advanced image processing algorithms can automatically detect and analyze vascular structures, reduce human errors, and enhance the diagnostic capabilities of DSA technology.

[0003] In the existing technology, medical personnel will obtain cerebrovascular images through DSA technology, locate the patient's cerebrovascular lesions, and outline parameters such as the tumor position, length, diameter, etc. This method is highly subjective, and due to the patient's own breathing behavior, it may cause displacement interference, which further increases the ambiguity and complexity of the cerebrovascular image itself, and ultimately makes it impossible for relevant personnel to accurately locate the position of cerebrovascular lesions. Summary of the invention

[0004] In order to solve the technical problem that the patient's breathing behavior may cause displacement interference, thereby further increasing the ambiguity and complexity of the cerebrovascular image itself, and ultimately causing the relevant personnel to be unable to accurately locate the position of cerebrovascular lesions, the purpose of the present invention is to provide a cerebrovascular lesion positioning method and system based on image recognition, and the technical scheme adopted is as follows: a cerebrovascular lesion positioning method based on image recognition, the method comprising: acquiring a cerebrovascular image of a patient; obtaining a cerebrovascular area in the cerebrovascular image; obtaining all cerebrovascular edge pixel points based on the distribution of the cerebrovascular area in the cerebrovascular image; establishing a preset number of preset sliding windows with all cerebrovascular edge pixel points as starting points, and placing all preset sliding windows The cerebrovascular region is traversed along the extension direction of the cerebrovascular vessel; all sliding window regions passed by the preset sliding window when moving are obtained, and the sliding window region where any preset sliding window is located is used as the reference sliding window region; the cerebrovascular expansion degree in the reference sliding window region is obtained according to the pixel point distribution of the cerebrovascular region; the expansion coefficient of the reference sliding window region is obtained according to the pixel point distribution of the cerebrovascular region in the reference sliding window region and the cerebrovascular expansion degree; the abnormal coefficient of the reference sliding window region is obtained according to the difference in expansion coefficients between the reference sliding window region and other sliding window regions, and the distribution of expansion coefficients of other sliding window regions; all sliding window regions are screened according to the abnormal coefficients to obtain all abnormal cerebrovascular regions; and cerebrovascular lesions are located according to the abnormal cerebrovascular regions.

[0005] Furthermore, the method for obtaining the cerebrovascular edge pixel points includes: using a skeletonization algorithm to simplify the cerebrovascular area in the cerebrovascular image into a cerebrovascular centerline; using all intersections of the cerebrovascular centerline and the boundary of the cerebrovascular image as centerline boundary points of the cerebrovascular centerline; and using all pixel points of the cerebrovascular area corresponding to the centerline boundary points as cerebrovascular edge pixel points.

[0006] Furthermore, the method for obtaining the degree of cerebral vascular expansion in the reference sliding window area includes: establishing a preset number of preset sliding windows with consistent width starting from the cerebral vascular edge pixel points; all preset sliding windows cover all cerebral vascular edge pixel points, and all preset sliding windows are adjacent to each other; two preset sliding windows intersecting with the boundary of the cerebral vascular area among all the preset sliding windows are used as sliding windows to be analyzed, and the absolute value of the slope of the straight line equation between each intersection of the sliding window to be analyzed and the boundary of the cerebral vascular area is calculated as the degree of cerebral vascular expansion in the reference sliding window area.

[0007] Furthermore, the method for obtaining the expansion coefficient of the reference sliding window area includes: taking a sliding window other than the sliding window to be analyzed as a main sliding window, and taking the area where the main sliding window is located as a main sliding window area; obtaining the intersection of the straight line equations of two sliding window areas to be analyzed in the reference sliding window area according to the degree of cerebral vascular expansion in the reference sliding window area; taking the distance between the central pixel point of the sliding window area to be analyzed that intersects with the upper boundary of the cerebral vascular area and the intersection of the straight line equation as a first distance; taking the distance between the central pixel point of the sliding window area to be analyzed that intersects with the lower boundary of the cerebral vascular area and the intersection of the straight line equation as a second distance; and obtaining the expansion coefficient according to an expansion coefficient calculation formula, and the expansion coefficient calculation formula is as follows: In the formula, represents the expansion coefficient of the reference sliding window area; Indicates the number of pixels in the sliding window area to be analyzed that intersects with the upper boundary of the cerebrovascular area; Represents the number of pixels in each main sliding window area; Represents the number of pixels in the sliding window area to be analyzed that intersects with the lower boundary of the cerebrovascular area, where ; Indicates the first The gray value of each pixel; In each main sliding window area, The gray value of each pixel; The first sliding window area to be analyzed intersecting with the lower boundary of the cerebrovascular area The gray value of each pixel; represents the first distance; Indicates the second distance.

[0008] Furthermore, the method for obtaining the abnormal coefficient of the reference sliding window area includes: screening each sliding window area according to the expansion coefficient of each sliding window area to obtain all normal sliding window areas; obtaining the abnormal coefficient according to the abnormal coefficient calculation formula, and the abnormal coefficient calculation formula is as follows: In the formula, represents the anomaly coefficient of the reference sliding window area; represents the expansion coefficient of the reference sliding window area; Indicates the number of normal sliding window areas outside the reference sliding window area; The reference window area is outside the The expansion coefficient of the normal sliding window area; Indicates a preset first threshold; It represents the standard deviation of the expansion coefficient of the normal sliding window area outside the reference sliding window area; Represents Iverson brackets, where if the condition in the Iverson brackets is met, the value is 1, and if the condition is not met, the value is 0.

[0009] Furthermore, the method for acquiring the normal sliding window area includes: taking the sliding window area whose expansion coefficient is less than a preset second threshold as the normal sliding window area.

[0010] Furthermore, all sliding window areas are screened according to the abnormal coefficient to obtain all abnormal cerebrovascular areas, including: marking the sliding window areas with abnormal coefficients greater than 0 as the first abnormal sliding window areas; and performing contour fitting on the cerebrovascular areas in all the first abnormal sliding window areas to obtain abnormal cerebrovascular areas.

[0011] A cerebrovascular lesion localization system based on image recognition, the system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of the above-mentioned cerebrovascular lesion localization method based on image recognition are implemented.

[0012] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned method for locating cerebrovascular lesions based on image recognition are implemented.

[0013] The present invention has the following beneficial effects: the present invention obtains a cerebrovascular image of a patient; obtains a cerebrovascular region in the cerebrovascular image; since cerebrovascular lesions are usually manifested as cerebrovascular tumors, in order to avoid missing the location of cerebrovascular tumors, the cerebrovascular edge pixel points of the cerebrovascular image are firstly extracted, and a preset number of preset sliding windows are established with all cerebrovascular edge pixel points as starting points to traverse the cerebrovascular region; since the characteristic of the location where the tumor occurs in the cerebrovascular region is that the part of the cerebrovascular region is expanded compared with other cerebrovascular regions, the degree of cerebrovascular expansion in the cerebrovascular region is analyzed; since when the preset sliding window traverses the cerebrovascular region, if the preset sliding window passes through the cerebrovascular region, the cerebrovascular region is expanded. The position of the cerebrovascular lesion is passed. At this time, the number of pixel points in the cerebrovascular area in the preset sliding window is more than the normal position of the preset sliding window passing through the cerebrovascular. Therefore, in addition to the degree of cerebrovascular expansion, the number of pixel points in the cerebrovascular area in the preset sliding window area should also be combined with the number of pixel points in the cerebrovascular area in the preset sliding window area, and the expansion coefficient of the reference sliding window area should be obtained by analysis; each time the preset sliding window slides, the expansion coefficient corresponding to this sliding can be obtained, and the sliding window area with abnormalities can be analyzed through the change of the expansion coefficient to obtain the abnormal coefficient of each sliding window area; all sliding window areas are screened according to the abnormal coefficient to obtain all abnormal cerebrovascular areas; cerebrovascular lesions are located according to the abnormal cerebrovascular areas. The present invention avoids the errors that may be caused by the ambiguity and complexity of the cerebrovascular image itself, thereby enabling relevant personnel to accurately locate the position of cerebrovascular lesions. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] 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.

[0015] Figure 1 A flow chart of a method for locating cerebrovascular lesions based on image recognition provided by one embodiment of the present invention; Figure 2 A flow chart of a cerebrovascular lesion localization system based on image recognition provided by one embodiment of the present invention; Figure 3 A process diagram of traversing a cerebral blood vessel region using a sliding window provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0016] 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 a method and system for locating cerebrovascular lesions based on image recognition proposed by the present invention, its specific implementation, structure, features and effects, in conjunction 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.

[0017] 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.

[0018] The following is a detailed description of a method and system for locating cerebrovascular lesions based on image recognition provided by the present invention in conjunction with the accompanying drawings.

[0019] See also Figure 1 , which shows a method for locating cerebrovascular lesions based on image recognition provided by an embodiment of the present invention, the method comprising: step S1: acquiring a cerebrovascular image of a patient.

[0020] The embodiment of the present invention is mainly applied to the scene of locating the position of cerebrovascular lesions. In one embodiment of the present invention, the patient's cerebrovascular images are collected by professional DSA equipment. The specific collection process is a technical means well known to those skilled in the art and will not be described in detail here.

[0021] In one embodiment of the present invention, the cerebrovascular image is grayed, and then the grayed cerebrovascular image is denoised to obtain a final cerebrovascular image. The graying operation and denoising method are both well-known technical means to those skilled in the art. In the embodiment of the present invention, the NLM algorithm is used to denoise the cerebrovascular image, which is not limited or elaborated herein.

[0022] Step S2: obtaining a cerebrovascular region in a cerebrovascular image; obtaining all cerebrovascular edge pixel points according to the cerebrovascular region distribution in the cerebrovascular image; establishing a preset number of preset sliding windows with all cerebrovascular edge pixel points as starting points, and traversing the cerebrovascular region along the extension direction of the cerebrovascular vessel with all preset sliding windows; obtaining all sliding window regions passed by the preset sliding windows when the preset sliding windows move, and taking the sliding window region where any preset sliding window is located as a reference sliding window region; obtaining the degree of cerebrovascular expansion in the reference sliding window region according to the pixel point distribution of the cerebrovascular region; obtaining the expansion coefficient of the reference sliding window region according to the pixel point distribution of the cerebrovascular region in the reference sliding window region and the degree of cerebrovascular expansion; obtaining the abnormal coefficient of the reference sliding window region according to the difference in expansion coefficients between the reference sliding window region and other sliding window regions, and the distribution of expansion coefficients of other sliding window regions.

[0023] In one embodiment of the present invention, all cerebrovascular regions in a cerebrovascular image are obtained according to an image segmentation algorithm, wherein the image segmentation algorithm is a technical means well known to those skilled in the art and is not limited or elaborated herein.

[0024] Since cerebrovascular lesions are usually manifested as cerebrovascular tumors, in order to avoid missing the location of cerebrovascular tumors, preferably, in one embodiment of the present invention, a sliding window is established for traversal of the cerebrovascular region, specifically including: first, extracting cerebrovascular edge pixel points of the cerebrovascular image.

[0025] Preferably, in one embodiment of the present invention, a method for acquiring cerebral blood vessel edge pixel points includes: using a skeletonization algorithm to simplify the cerebral blood vessel region in a cerebral blood vessel image into a cerebral blood vessel centerline. It should be noted that the skeletonization algorithm is a technical means well known to those skilled in the art and will not be described in detail here.

[0026] All intersection points between the cerebrovascular centerline and the cerebrovascular image boundary are taken as the centerline boundary points of the cerebrovascular centerline; and the pixel points of the cerebrovascular area corresponding to all the centerline boundary points are taken as the cerebrovascular edge pixel points.

[0027] Secondly, a preset number of preset sliding windows are established with all cerebrovascular edge pixel points as starting points. The specific process includes: a preset number of preset sliding windows with consistent widths are established with the cerebrovascular edge pixel points as starting points, and all the preset sliding windows cover all cerebrovascular edge pixel points, and all the preset sliding windows are adjacent to each other.

[0028] Traverse the cerebrovascular area along the extension direction of the cerebrovascular all preset sliding windows. For details, please refer to Figure 3 , provides a process diagram of sliding window traversing the cerebrovascular region. In one embodiment of the present invention, the preset number is set to 3. At this time, the 3 preset sliding windows are set to: , all sliding window areas passed by different preset sliding windows when they move are obtained. It should be noted that in other embodiments of the present invention, the preset number and the size of the preset sliding window can be set by oneself.

[0029] Since the location where the cerebrovascular region undergoes tumor change is characterized by the expansion of the cerebrovascular region compared to other cerebrovascular regions, in the embodiment of the present invention, the degree of expansion of the cerebrovascular region is analyzed.

[0030] Preferably, in one embodiment of the present invention, the method for obtaining the degree of cerebral vascular expansion in the reference sliding window region includes: taking two preset sliding windows intersecting with the boundary of the cerebral vascular region among all preset sliding windows as sliding windows to be analyzed, and calculating the absolute value of the slope of the straight line equation between the intersection of each sliding window to be analyzed and the boundary of the cerebral vascular region as the degree of cerebral vascular expansion in the reference sliding window region, wherein the greater the absolute value of the slope of the straight line equation, the greater the degree of expansion of the cerebral vascular region in the corresponding sliding window to be analyzed. It should be noted that the method for obtaining the slope of the straight line equation is a technical means well known to those skilled in the art and will not be described in detail here.

[0031] Since when the preset sliding window passes through the position of cerebrovascular lesions when traversing the cerebrovascular region, the number of cerebrovascular region pixel points in the preset sliding window is greater than when the preset sliding window passes through the normal position of the cerebrovascular region, in the embodiment of the present invention, in addition to the degree of cerebrovascular expansion, the number of cerebrovascular region pixel points in the preset sliding window region should also be considered. Therefore, the expansion coefficient of the reference sliding window region is obtained based on the distribution of cerebrovascular region pixel points and the degree of cerebrovascular expansion in the reference sliding window region.

[0032] Preferably, in one embodiment of the present invention, the method for obtaining the expansion coefficient of the reference sliding window region includes: taking the sliding windows other than the sliding window to be analyzed as the main sliding window, taking the region where the main sliding window is located as the main sliding window region, Figure 3 As shown, the pixel points in the main sliding window area are all pixel points in the cerebrovascular area; according to the cerebrovascular expansion degree of the reference sliding window area, the intersection point of the straight line equation of the two sliding window areas to be analyzed in the reference sliding window area is obtained, wherein, since the cerebrovascular expansion degree is the slope of the straight line equation between the intersection point of each sliding window to be analyzed and the boundary of the cerebrovascular area, the calculation process is a technical means well known to those skilled in the art and will not be described in detail here.

[0033] The distance between the central pixel point of the sliding window area to be analyzed that intersects the upper boundary of the cerebrovascular area and the intersection point of the straight line equation is taken as the first distance; the distance between the central pixel point of the sliding window area to be analyzed that intersects the lower boundary of the cerebrovascular area and the intersection point of the straight line equation is taken as the second distance; in one embodiment of the present invention, the calculation formulas of the first distance and the second distance are as follows: , In the formula, represents the first distance; Represents the abscissa of the intersection point between the two straight line equations formed by the intersection points of the two sliding windows to be analyzed and the cerebrovascular region; The ordinate of the intersection point between the two straight line equations formed by the intersection points of the two sliding windows to be analyzed and the cerebrovascular region; Represents the horizontal coordinate of the center pixel of the first sliding window area to be analyzed; The vertical coordinate of the center pixel of the first sliding window area to be analyzed; represents the second distance; Represents the horizontal coordinate of the center pixel of the second sliding window area to be analyzed; Indicates the vertical coordinate of the center pixel of the second sliding window area to be analyzed.

[0034] The expansion coefficient is obtained according to the expansion coefficient calculation formula. The expansion coefficient calculation formula is as follows: In the formula, represents the expansion coefficient of the reference sliding window area; Indicates the number of pixels in the sliding window area to be analyzed that intersects with the upper boundary of the cerebrovascular area; Represents the number of pixels in each main sliding window area; Represents the number of pixels in the sliding window area to be analyzed that intersects with the lower boundary of the cerebrovascular area, where ; Indicates the first The gray value of each pixel; In each main sliding window area, The gray value of each pixel; The first sliding window area to be analyzed intersecting with the lower boundary of the cerebrovascular area The gray value of each pixel; represents the first distance; Indicates the second distance.

[0035] In the expansion coefficient calculation formula, the mean difference between the grayscale values ​​of the pixels in the sliding window area to be analyzed and the grayscale values ​​of the pixels in the cerebrovascular area is The smaller it is, the more likely the sliding window area to be analyzed is to pass through the area of ​​cerebral vascular expansion, and the larger the expansion coefficient of the corresponding reference sliding window area is. If the preset sliding window passes through the normal blood vessel area, the positional relationship between the two straight line equations is approximately parallel. However, if the preset sliding window passes through the area of ​​cerebral vascular lesions, the two straight line equations will intersect, and the greater the expansion degree, the closer the intersection of the two straight line equations is to the central pixel points of the two sliding window areas to be analyzed. At this time, the sum of the first distance and the second distance is The smaller it is, the larger the expansion coefficient of the corresponding reference sliding window area is.

[0036] Each time the preset sliding window slides, the expansion coefficient corresponding to this sliding can be obtained. The sliding window area with abnormalities can be analyzed by the change of the expansion coefficient to obtain the abnormal coefficient of each sliding window area. Therefore, in the embodiment of the present invention, the abnormal coefficient of the reference sliding window area is obtained according to the difference in expansion coefficients between the reference sliding window area and other sliding window areas, and the distribution of expansion coefficients of other sliding window areas.

[0037] Preferably, in one embodiment of the present invention, a method for obtaining an abnormal coefficient of a reference sliding window area includes: taking a sliding window area whose expansion coefficient is less than a preset second threshold as a normal sliding window area. In one embodiment of the present invention, the preset second threshold is set to 0.2. It should be noted that, in other embodiments of the present invention, the preset second threshold can be set by itself and is not limited here.

[0038] The abnormal coefficient is obtained according to the abnormal coefficient calculation formula. The abnormal coefficient calculation formula is as follows: In the formula, represents the anomaly coefficient of the reference sliding window area; represents the expansion coefficient of the reference sliding window area; Indicates the number of normal sliding window areas outside the reference sliding window area; The reference window area is outside the The expansion coefficient of the normal sliding window area; Indicates a preset first threshold; It represents the standard deviation of the expansion coefficient of the normal sliding window area outside the reference sliding window area; Represents Iverson brackets, where if the condition in the Iverson brackets is met, the value is 1, and if the condition is not met, the value is 0.

[0039] In the calculation formula of the abnormal coefficient, by presetting the first threshold Screening the abnormal area, the preset first threshold is adjustable, and The larger the value of is, the looser the screening of abnormal sliding window areas is. Some of the result values ​​exceed At this time , then we get The value of will not be , the sliding window area can be considered as an abnormal sliding window area, and further analysis is performed on the sliding window area; if the difference between the expansion coefficient of the reference sliding window area and the mean expansion coefficient of all normal sliding window areas is The larger the value, the more likely it is that there is cerebrovascular abnormality in the reference sliding window area; the standard deviation of the expansion coefficient of the normal sliding window area outside the reference sliding window area The smaller it is, the more stable the change in the expansion coefficient of the normal sliding window area is. At this time, the greater the difference between the expansion coefficient of the reference sliding window area and the normal sliding window area, the more likely it is that there is cerebrovascular abnormality in the reference sliding window area, and the greater the abnormal coefficient of the reference sliding window area is.

[0040] Step S3: Screen all sliding window areas according to the abnormal coefficients to obtain all abnormal cerebrovascular areas; locate cerebrovascular lesions according to the abnormal cerebrovascular areas.

[0041] Preferably, in one embodiment of the present invention, all sliding window areas are screened according to the abnormal coefficient to obtain all abnormal cerebrovascular areas, including: marking the sliding window areas with abnormal coefficients greater than 0 as the first abnormal sliding window areas; performing contour fitting on the cerebrovascular areas in all the first abnormal sliding window areas to obtain abnormal cerebrovascular areas.

[0042] The location of cerebrovascular lesions is localized based on the abnormal cerebrovascular area.

[0043] In summary, the patient's cerebrovascular image is obtained; the cerebrovascular area in the cerebrovascular image is obtained; according to the distribution of the cerebrovascular area in the cerebrovascular image, all cerebrovascular edge pixel points are obtained; a preset number of preset sliding windows are established with all cerebrovascular edge pixel points as starting points, and all preset sliding windows are traversed along the extension direction of the cerebrovascular area; all sliding window areas passed by the preset sliding window when moving are obtained, and the sliding window area where any preset sliding window is located is used as the reference sliding window area; according to the pixel point distribution of the cerebrovascular area, the degree of cerebrovascular expansion in the reference sliding window area is obtained; according to the pixel point distribution of the cerebrovascular area in the reference sliding window area and the degree of cerebrovascular expansion, the expansion coefficient of the reference sliding window area is obtained; according to the difference in expansion coefficients between the reference sliding window area and other sliding window areas, and the distribution of expansion coefficients of other sliding window areas, the abnormal coefficient of the reference sliding window area is obtained; according to the abnormal coefficient, all sliding window areas are screened to obtain all abnormal cerebrovascular areas; and cerebrovascular lesions are located according to the abnormal cerebrovascular areas.

[0044] An embodiment of the present invention provides a cerebrovascular lesion positioning system based on image recognition, specifically Figure 2As shown, the system includes a memory, a processor and a computer program, wherein the memory is used to store the corresponding computer program, the processor is used to run the corresponding computer program, and when the computer program runs in the processor, the method described in steps S1-S3 can be implemented. The system includes: an image acquisition module S101, a cerebrovascular area analysis module S102, and a cerebrovascular lesion positioning module S103.

[0045] An image acquisition module is used to obtain a cerebrovascular image of a patient; a cerebrovascular region analysis module is used to obtain a cerebrovascular region in a cerebrovascular image; according to the distribution of the cerebrovascular region in the cerebrovascular image, all cerebrovascular edge pixel points are obtained; a preset number of preset sliding windows are established with all cerebrovascular edge pixel points as starting points, and all preset sliding windows are used to traverse the cerebrovascular region along the extension direction of the cerebrovascular; all sliding window regions passed by the preset sliding window when the preset sliding window moves are obtained, and the sliding window region where any preset sliding window is located is used as a reference sliding window region; according to the pixel point distribution of the cerebrovascular region, the expansion degree of the cerebrovascular in the reference sliding window region is obtained; according to the pixel point distribution of the cerebrovascular region in the reference sliding window region and the expansion degree of the cerebrovascular, the expansion coefficient of the reference sliding window region is obtained; according to the difference in expansion coefficients between the reference sliding window region and other sliding window regions, and the distribution of expansion coefficients of other sliding window regions, the abnormal coefficient of the reference sliding window region is obtained; a cerebrovascular lesion positioning module is used to screen all sliding window regions according to the abnormal coefficients, obtain all abnormal cerebrovascular regions; and locate cerebrovascular lesions according to the abnormal cerebrovascular regions.

[0046] The third purpose of an embodiment of the present invention is to provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned method for locating cerebrovascular lesions based on image recognition when executing the computer program.

[0047] The fourth objective of an embodiment of the present invention is to provide a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned method for locating cerebrovascular lesions based on image recognition are implemented.

[0048] 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.

[0049] 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 method for locating cerebrovascular lesions based on image recognition, characterized in that: The method comprises: obtaining a cerebrovascular image of a patient; obtaining a cerebrovascular region in the cerebrovascular image; obtaining all cerebrovascular edge pixel points according to the distribution of the cerebrovascular region in the cerebrovascular image; establishing a preset number of preset sliding windows with all cerebrovascular edge pixel points as starting points, and traversing the cerebrovascular region along the extension direction of the cerebrovascular with all the preset sliding windows; obtaining all sliding window regions passed by the preset sliding windows when the preset sliding windows move, and taking the sliding window region where any preset sliding window is located as a reference sliding window region; obtaining the cerebrovascular expansion degree in the reference sliding window region according to the pixel point distribution of the cerebrovascular region; obtaining the expansion coefficient of the reference sliding window region according to the pixel point distribution of the cerebrovascular region in the reference sliding window region and the cerebrovascular expansion degree; obtaining the abnormal coefficient of the reference sliding window region according to the difference in expansion coefficients between the reference sliding window region and other sliding window regions, and the distribution of expansion coefficients of other sliding window regions; screening all sliding window regions according to the abnormal coefficients to obtain all abnormal cerebrovascular regions; locating cerebrovascular lesions according to the abnormal cerebrovascular regions; the method for obtaining the cerebrovascular expansion degree in the reference sliding window region comprises: A preset number of preset sliding windows with the same width are established starting from the pixel points at the edge of the cerebrovascular vessel; all the preset sliding windows cover all the pixel points at the edge of the cerebrovascular vessel, and all the preset sliding windows are adjacent to each other; two preset sliding windows intersecting with the boundary of the cerebrovascular region among all the preset sliding windows are taken as sliding windows to be analyzed, and the absolute value of the slope of the straight line equation between the intersection points of each of the sliding windows to be analyzed and the boundary of the cerebrovascular region is calculated as the degree of cerebrovascular expansion in the reference sliding window region; the method for obtaining the expansion coefficient of the reference sliding window region includes: taking the sliding windows other than the sliding windows to be analyzed as the main sliding windows, and taking the area where the main sliding windows are located as the main sliding window area; obtaining the intersection point of the straight line equation of two sliding window areas to be analyzed in the reference sliding window area according to the degree of cerebrovascular expansion in the reference sliding window area; taking the distance between the central pixel point of the sliding window area to be analyzed intersecting with the upper boundary of the cerebrovascular region and the intersection point of the straight line equation as the first distance; taking the distance between the central pixel point of the sliding window area to be analyzed intersecting with the lower boundary of the cerebrovascular region and the intersection point of the straight line equation as the second distance; obtaining the expansion coefficient according to the expansion coefficient calculation formula, and the expansion coefficient calculation formula is as follows: In the formula, represents the expansion coefficient of the reference sliding window area; Indicates the number of pixels in the sliding window area to be analyzed that intersects with the upper boundary of the cerebrovascular area; Represents the number of pixels in each main sliding window area; Represents the number of pixels in the sliding window area to be analyzed that intersects with the lower boundary of the cerebrovascular area, where ; Indicates the first The gray value of each pixel; In each main sliding window area, The gray value of each pixel; The first sliding window area to be analyzed intersecting with the lower boundary of the cerebrovascular area The gray value of each pixel; represents the first distance; represents the second distance; the method for obtaining the abnormal coefficient of the reference sliding window area includes: screening each sliding window area according to the expansion coefficient of each sliding window area to obtain all normal sliding window areas; obtaining the abnormal coefficient according to the abnormal coefficient calculation formula, and the abnormal coefficient calculation formula is as follows: In the formula, represents the anomaly coefficient of the reference sliding window area; represents the expansion coefficient of the reference sliding window area; Indicates the number of normal sliding window areas outside the reference sliding window area; The reference window area is outside the The expansion coefficient of the normal sliding window area; Indicates a preset first threshold; It represents the standard deviation of the expansion coefficient of the normal sliding window area outside the reference sliding window area; Represents Iverson brackets, where if the condition in the Iverson brackets is met, the value is 1, and if the condition is not met, the value is 0.

2. The method for locating cerebrovascular lesions based on image recognition according to claim 1, characterized in that: The method for acquiring cerebrovascular edge pixel points includes: using a skeletonization algorithm to simplify the cerebrovascular region in the cerebrovascular image into a cerebrovascular centerline; using all intersections of the cerebrovascular centerline and the boundary of the cerebrovascular image as centerline boundary points of the cerebrovascular centerline; and using pixel points of the cerebrovascular region corresponding to all the centerline boundary points as cerebrovascular edge pixel points.

3. The method for locating cerebrovascular lesions based on image recognition according to claim 1, characterized in that: The method for acquiring a normal sliding window region includes: taking a sliding window region whose expansion coefficient is less than a preset second threshold as a normal sliding window region.

4. The method for locating cerebrovascular lesions based on image recognition according to claim 1, characterized in that: All sliding window areas are screened according to the abnormal coefficient to obtain all abnormal cerebrovascular areas, including: marking the sliding window areas with abnormal coefficients greater than 0 as the first abnormal sliding window areas; and performing contour fitting on the cerebrovascular areas in all the first abnormal sliding window areas to obtain the abnormal cerebrovascular areas.

5. A cerebrovascular lesion localization system based on image recognition, the system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of a method for locating cerebrovascular lesions based on image recognition as described in any one of claims 1 to 4 are implemented.

6. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of a method for locating cerebrovascular lesions based on image recognition as claimed in any one of claims 1 to 4 are implemented.

Images