A method for assessing vascular stenosis based on cardiac angiogram images

Abstract

The present application relates to the technical field of image data processing, in particular to a kind of blood vessel stenosis evaluation method based on cardiac angiography image, comprising: obtaining the suspected blood vessel contour region in the cardiac angiography image of patient, segmentation into several suspected blood vessel branch region, to filter out heart blood vessel branch region, according to the change of blood vessel width in heart blood vessel branch region, first type heart blood vessel stenosis region is segmented out, according to the similarity between the end points of the broken skeleton line between heart blood vessel branch region, determine the second type heart blood vessel stenosis region of broken place, linear transformation is carried out to the heart blood vessel stenosis region in cardiac angiography image, obtain cardiac angiography enhancement image, quantitative analysis is carried out to cardiac angiography enhancement image, obtain the stenosis degree of heart blood vessel.The present application improves the accuracy of blood vessel stenosis evaluation by contrast enhancement to unclear heart blood vessel stenosis region.

Description

Technical Field

[0001] This invention relates to the field of image data processing technology, and more specifically to a method for assessing vascular stenosis based on cardiac angiography images. Background Technology

[0002] Angiography allows doctors to clearly observe the morphology and structure of coronary arteries and other blood vessels, thereby detecting lesions such as stenosis and blockage. Vascular stenosis can lead to insufficient blood supply to the heart muscle, increasing the risk of heart attack; therefore, timely and accurate assessment of the degree of stenosis is crucial for developing personalized treatment plans. Through angiography images, doctors can assess the location and extent of stenosis, as well as the presence of multiple lesions, thus determining whether interventional treatment, stent implantation, or surgery is necessary. Therefore, vascular stenosis assessment is an indispensable part of the diagnosis and treatment of heart disease.

[0003] Existing problem: Quantitative angiography is a computer-aided technique for analyzing angiographic images, mainly used to quantitatively assess the degree of vascular stenosis. However, due to the inherently low signal-to-noise ratio of X-ray angiography, reduced blood flow in the stenotic area may lead to uneven distribution of contrast agent within the blood vessel, making the image of that area appear blurry or unclear, which may result in inaccurate assessment of vascular stenosis. Summary of the Invention

[0004] This invention provides a method for assessing vascular stenosis based on cardiac angiography images to address existing problems.

[0005] The present invention provides a method for assessing vascular stenosis based on cardiac angiography images, which employs the following technical solution:

[0006] One embodiment of the present invention provides a method for assessing vascular stenosis based on cardiac angiography images, the method comprising the following steps:

[0007] Obtain cardiac angiography images of the patient;

[0008] The suspected vascular contour region in the cardiac angiography image is obtained, and the region is segmented by the branch points on the skeleton line of the suspected vascular contour region. The segmented skeleton line is then used to divide the suspected vascular contour region into several suspected vascular branch regions.

[0009] Based on the shape of the suspected vascular branching region, cardiac vascular branching regions are screened out; based on the changes in the vascular width within the cardiac vascular branching regions, the first type of cardiac vascular stenosis region is segmented.

[0010] Based on the similarity between the endpoints of the broken skeleton lines between the branch regions of the heart vessels, the second type of heart vessel stenosis region at the break point is identified.

[0011] Linear transformation is performed on the narrowed areas of the heart vessels in the cardiac angiography images to obtain enhanced cardiac angiography images; quantitative analysis is then performed on the enhanced cardiac angiography images to obtain the degree of narrowing of the heart vessels.

[0012] Furthermore, the specific steps for screening cardiac vascular branch regions based on the shape of suspected vascular branch regions are as follows:

[0013] The distance between the normal of each pixel on the skeleton line of the suspected blood vessel branch region and the intersection point of the boundary of the suspected blood vessel branch region is taken as the width of each pixel on the skeleton line;

[0014] Based on the width of each pixel on the skeleton line, the likelihood of cardiac blood vessels in the suspected vascular branch area is determined.

[0015] Based on the probability of the cardiac blood vessels, the branch regions of the cardiac blood vessels are screened out.

[0016] Furthermore, the specific steps for determining the likelihood of cardiac vessels in a suspected vascular branch region based on the width of each pixel on the skeleton line are as follows:

[0017] The length of the skeleton line of the suspected vascular branch region and the mean of the width of all pixels on the skeleton line are obtained. The ratio of the length to the mean is calculated and recorded as the first ratio. Then, the variance of the gradient values ​​of all pixels on the boundary of the suspected vascular branch region is obtained. The normalized value of the ratio of the first ratio to the variance is used as the probability of the cardiac blood vessels in the suspected vascular branch region.

[0018] Furthermore, the specific steps for screening out cardiac vessel branch regions based on the probability of the cardiac vessels are as follows:

[0019] Suspected vascular branch regions where the probability of a cardiac vascular branch is greater than a preset vascular threshold are denoted as cardiac vascular branch regions.

[0020] Furthermore, the specific steps for segmenting the first type of coronary artery stenosis region based on the changes in vessel width in the branch regions of the coronary arteries are as follows:

[0021] A preset length threshold L is used to divide the skeleton line of the cardiac blood vessel branch area into several skeleton line segments of length L.

[0022] The narrowness of a skeleton segment is determined by the difference in the width between pixels on the skeleton segment.

[0023] Skeleton segments with a narrowing probability greater than a preset narrowing threshold are denoted as narrow skeleton segments;

[0024] The cardiac vascular branching region is divided into several sub-regions corresponding to skeletal segments. The sub-regions corresponding to the narrow skeletal segments are denoted as the first type of cardiac vascular stenosis region.

[0025] Furthermore, the specific steps for determining the narrowness probability of a skeleton line segment based on the difference in width between pixels on the skeleton line segment are as follows:

[0026] For any skeleton line segment, calculate the average width of all pixels, and use the normalized value of the average of the absolute values ​​of the differences between the widths of all pixels and the average width as the narrowness probability of the arbitrary skeleton line segment.

[0027] Furthermore, the specific steps for determining the second type of coronary artery stenosis region at the break point based on the similarity between the endpoints of the broken skeletal lines between the branch regions of the coronary blood vessels are as follows:

[0028] Among the endpoints of the skeleton lines of all suspected vascular contour areas, the endpoints of the skeleton lines that are also branches of the heart and blood vessels are recorded as suspected stenosis endpoints.

[0029] The suspected narrow endpoint that is closest to the h-th suspected narrow endpoint is denoted as the reference suspected narrow endpoint;

[0030] Based on the distance and width between the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint, determine the probability that the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are the same blood vessel;

[0031] With a preset threshold N, the skeleton line of the cardiac vascular branch region where the suspected stenosis endpoint is located is traversed pixel by pixel starting from the suspected stenosis endpoint. The least squares method is used to fit a straight line to the first N traversed pixels to obtain the fitted straight line of the suspected stenosis endpoint.

[0032] Based on the angle between the fitted straight lines of the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint, and the probability that they are the same blood vessel, the final probability that the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are the same blood vessel is determined.

[0033] Based on the probability that the stenosis is due to the same blood vessel, a second type of cardiac vascular stenosis region is obtained.

[0034] Furthermore, the specific steps for determining the probability that the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are the same blood vessel based on the distance and width between the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are as follows:

[0035] Calculate the absolute value of the difference between the width of the h-th suspected stenosis endpoint and the width of the reference suspected stenosis endpoint. Take the inversely proportional normalized value of the product of the distance between the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint and the absolute value of the difference as the probability that the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are the same blood vessel.

[0036] Furthermore, the specific steps involved in determining the final probability that the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are the same vessel, based on the angle between the fitted straight lines of the h-th suspected stenosis endpoint and the probability that they are the same vessel, are as follows:

[0037] Obtain the minimum angle between the fitted line of the h-th suspected stenosis endpoint and the fitted line of the reference suspected stenosis endpoint. Use the normalized value of the ratio of the probability that the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are the same blood vessel to the minimum angle as the final probability that the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are the same blood vessel.

[0038] Furthermore, the specific steps for obtaining the second type of coronary artery stenosis region based on the final probability of it being the same blood vessel are as follows:

[0039] When the final probability that the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are the same blood vessel is greater than the preset matching threshold, the maximum width between the width of the h-th suspected stenosis endpoint and the width of the reference suspected stenosis endpoint is obtained. Starting from the straight line segment connecting the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint, half of the maximum width is extended to both sides from the perpendicular direction of the connecting straight line segment to construct a rectangular region as the second type of cardiovascular stenosis region.

[0040] The beneficial effects of the technical solution of the present invention are:

[0041] In this embodiment of the invention, a suspected vascular contour region is acquired from a patient's cardiac angiography image. The region is then segmented using branch points on the skeleton lines of this suspected vascular contour region. These segmented skeleton lines further divide the suspected vascular contour region into several suspected vascular branch regions, thereby identifying cardiac vascular branch regions. This selection of cardiac vascular branch regions based on vascular shape characteristics ensures the accuracy of subsequent vascular analysis. Based on the variation in vascular width within these branch regions, a first type of cardiac vascular stenosis region is segmented, thus identifying stenotic areas within the blood vessels. Then, based on the similarity between the endpoints of the broken skeleton lines between the cardiac vascular branch regions, a second type of cardiac vascular stenosis region is identified at the break points. This identifies stenotic areas that are not clearly visible in the image due to reduced blood flow, i.e., areas where the blood vessels are not broken, ensuring the accuracy of cardiac vascular stenosis region selection and consequently, the accuracy of subsequent vascular stenosis assessment. A linear transformation is performed on the cardiac vascular stenosis region in the cardiac angiography image to obtain an enhanced cardiac angiography image. Quantitative analysis of this enhanced image yields the degree of cardiac vascular stenosis. Thus, this invention improves the accuracy of vascular stenosis assessment by enhancing the contrast of unclear cardiac vascular stenosis regions. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a flowchart illustrating the steps of a method for assessing vascular stenosis based on cardiac angiography images according to the present invention.

[0044] Figure 2 A schematic diagram of a narrowed area in the heart's blood vessels;

[0045] Figure 3 This is a schematic diagram of a suspected narrow end. Detailed Implementation

[0046] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a method for assessing vascular stenosis based on cardiac angiography images proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0047] Unless otherwise defined, 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 pertains.

[0048] The following description, in conjunction with the accompanying drawings, details a specific scheme for a method for assessing vascular stenosis based on cardiac angiography images provided by the present invention.

[0049] Please see Figure 1 The diagram illustrates a flowchart of a method for assessing vascular stenosis based on cardiac angiography images, according to an embodiment of the present invention. The method includes the following steps:

[0050] Step S001: Obtain cardiac angiography images of the patient.

[0051] A digital subtraction angiography (DSA) machine is used to photograph the patient's heart to obtain cardiac angiography images.

[0052] It should be noted that in this embodiment, the cardiac angiography images have undergone mean filtering to remove noise from the cardiac angiography images. Mean filtering is a well-known technique, and the specific method will not be described here.

[0053] It should be further noted that during quantitative angiography, reduced blood flow at the site of vascular stenosis can lead to uneven distribution of contrast agent, causing the image in that area to become blurry or unclear, thus affecting the accuracy of the analysis of the degree of cardiac vascular stenosis. Therefore, in this embodiment, the characteristics of the vascular stenosis region are utilized to screen out the stenosis region and perform image enhancement to ensure the accuracy of vascular stenosis assessment. Specifically, the characteristics of the vascular stenosis region are as follows: when a cardiac blood vessel is blocked, the width of the blocked area narrows. When the blockage is severe, the same cardiac blood vessel may appear to be broken, with the two ends of the break having similar widths, being close together, and oriented in the same direction. Therefore, these characteristics can be used to screen out the stenosis region in the cardiac angiography image.

[0054] Step S002: Obtain the suspected vascular contour region in the cardiac angiography image, segment it according to the branch points on the skeleton line of the suspected vascular contour region, and use the segmented skeleton line to divide the suspected vascular contour region into several suspected vascular branch regions.

[0055] Preferably, in one embodiment of the present invention, the method for obtaining the suspected vascular branch region includes:

[0056] The Canny edge detection algorithm is used to process cardiac angiography images to obtain several suspected blood vessel edge pixels and the gradient value of each pixel. Then, the contour extraction algorithm is used to process all suspected blood vessel edge pixels to obtain several suspected blood vessel contour regions.

[0057] The Canny edge detection algorithm and contour extraction algorithm are both well-known technologies, and their specific methods will not be described here.

[0058] It should be noted that artifact regions often appear in cardiac vessels, and their presence can interfere with the screening and analysis of vascular regions. Vascular regions are generally elongated with smooth, regular edges, while artifact regions are more irregular in shape and have blurred edges. Therefore, this characteristic can be used to screen out several cardiac vascular branch regions.

[0059] Morphological refinement is used to calculate the skeleton lines of each suspected vascular contour region in the cardiac angiography image.

[0060] Using neighborhood analysis, the branch points on the skeleton line of each suspected blood vessel contour region are obtained.

[0061] The skeleton line of each suspected blood vessel contour region is divided at each branch point to obtain several skeleton line segments (segments with only two endpoints). Using the distance transformation method, each suspected blood vessel contour region is divided into several sub-regions based on the skeleton line segments, which are denoted as suspected blood vessel branch regions.

[0062] It should be noted that morphological thinning, neighborhood analysis, and distance transformation are all well-known techniques, and their specific methods will not be described here. Morphological thinning, during iterative thinning, applies morphological erosion to remove peripheral pixels of connected components while preserving the overall shape and topology of the connected components. Therefore, the endpoints of the skeleton lines are not boundary pixels of connected components, ensuring the acquisition of the width of the endpoint pixels on subsequent skeleton lines. Neighborhood analysis examines each skeleton line point and its eight neighboring skeleton line points. If a skeleton line point has three or more adjacent skeleton line points, it is identified as a branch point. Distance transformation calculates the distance from each pixel to the nearest skeleton line segment and assigns pixels to the nearest skeleton line segment based on these distances, forming the suspected vascular branch region corresponding to each skeleton line segment.

[0063] Step S003: Based on the shape of the suspected vascular branching region, screen out the cardiac vascular branching region; based on the changes in the vascular width in the cardiac vascular branching region, segment out the first type of cardiac vascular stenosis region.

[0064] Preferably, in one embodiment of the present invention, the method for obtaining the first type of coronary artery stenosis region includes:

[0065] For any suspected blood vessel branch region, obtain the normal of each pixel on the skeleton line of the suspected blood vessel branch region, and then obtain the first intersection point of each pixel on the skeleton line with the boundary of the suspected blood vessel branch region along both sides of its normal. The distance between these two first intersection points is taken as the width of each pixel on the skeleton line.

[0066] For any suspected vascular branch region, obtain the length of the skeleton line of the suspected vascular branch region and the mean of the width of all pixels on the skeleton line. Calculate the ratio of the length to the mean and record it as the first ratio. Then obtain the variance of the gradient values ​​of all pixels on the boundary of the suspected vascular branch region. The normalized value of the ratio of the first ratio to the variance is taken as the probability of the cardiac blood vessels in the suspected vascular branch region.

[0067] It should be noted that in this embodiment, the norm() linear normalization function is used to normalize the ratio of the first ratio to the variance, normalizing the data values ​​to the [0,1] interval. This is used as an example for description. The smoother the edge of the branch region, the smaller the variance of the gradient of all pixels on the boundary of the branch region should be, and the greater the probability that it is a cardiovascular system. Conversely, the larger the first ratio, the more elongated the branch region is, and the greater the probability that it is a cardiovascular system.

[0068] The preset blood vessel threshold is 0.7, and this will be used as an example for explanation.

[0069] Suspected vascular branch regions where the probability of a cardiac vascular branch is greater than a preset vascular threshold are denoted as cardiac vascular branch regions.

[0070] It should be noted that when a blood vessel in the heart becomes blocked, the width of certain areas of the vessel changes, becoming increasingly narrow, which restricts blood flow. This narrowing reduces the width of the blood vessel, limiting the ability of blood to flow to the heart, thereby affecting the heart's oxygen and nutrient supply. Based on this characteristic, several narrowed areas of the blood vessels in the heart were identified.

[0071] The preset length threshold L is 30, and this will be used as an example for explanation.

[0072] Divide the skeleton line of any cardiac vascular branch region into several skeleton line segments of length L.

[0073] It should be noted that: starting from one end of the skeleton line of the cardiac blood vessel branching region, the area is divided equally. If the length of the last skeleton line segment to the other end is less than 30, it will still be analyzed later.

[0074] On any skeleton segment equally divided by the skeleton line of any cardiac vascular branch region, calculate the average width of all pixels, then calculate the absolute value of the difference between the width of each pixel and the average width, and use the normalized value of the average of the absolute values ​​of the differences between the width of all pixels and the average width as the narrowing probability of the skeleton segment.

[0075] It should be noted that in this embodiment, the norm() linear normalization function is used to normalize the mean of the differences between the width of all pixels and the mean width, normalizing the data values ​​to the [0,1] interval. This is used as an example for description. Under normal circumstances, blood vessels should gradually decrease in size, meaning the difference between the width of each pixel and the mean width should be small, and the mean of the differences between the width of all pixels and the mean width should be close to 0. However, when a blood vessel becomes blocked, its width changes, first gradually narrowing and then slowly recovering. Therefore, a larger mean difference between the width of all pixels and the mean width indicates a more likely narrow segment of the blood vessel.

[0076] The preset narrowing threshold is 0.75, and this will be used as an example for explanation.

[0077] On the skeleton line of any cardiac vascular branch region, the skeleton line segment with a stenosis probability greater than a preset stenosis threshold is denoted as the stenosis skeleton line segment.

[0078] Then, using the distance transformation method, the cardiac vascular branch region is divided into several sub-regions based on the equally divided skeleton line segments of any given cardiac vascular branch region. Each skeleton line segment corresponds to one sub-region. The sub-region corresponding to the narrow skeleton line segment is denoted as the first type of cardiac vascular stenosis region. A schematic diagram of cardiac vascular stenosis regions is shown below. Figure 2 As shown, Figure 2 The area within the black circle represents a narrowed region of the heart's blood vessels.

[0079] Step S004: Based on the similarity between the endpoints of the broken skeleton lines between the branch regions of the heart vessels, determine the second type of heart vessel stenosis region at the break point.

[0080] It's important to note that when coronary arteries are severely blocked, it can appear as if the vessels are disconnected, even though they are not actually completely blocked. This occurs because severe blockage restricts blood flow, causing it to become more diffuse. Due to this diffuse nature, the blood flow is not clearly visible on imaging, potentially creating the illusion of gaps between vessels, which poses a challenge for doctors. Therefore, it's necessary to determine the possibility that two closely spaced vessel ends belong to the same vessel. If the ends of two vessels do belong to the same vessel, they typically have similar widths and are close together. Furthermore, the directions of these two ends usually align, tending to be parallel or nearly parallel.

[0081] Preferably, in one embodiment of the present invention, the method for obtaining the second type of coronary artery stenosis region includes:

[0082] Among the endpoints of the skeleton lines of all suspected vascular contour areas, the endpoints of the skeleton lines that are also branches of the heart and blood vessels are denoted as suspected stenosis endpoints.

[0083] Among them, the schematic diagram of the suspected narrow end is as follows: Figure 3 As shown, Figure 3 Black dots 1, 2, and 3 are the endpoints of the skeleton line of a suspected vascular contour area. If black dots 2 and 4 are the endpoints of the skeleton line of a cardiac vascular branch area in the suspected vascular contour area, then black dot 2 is the suspected stenosis endpoint.

[0084] Taking the h-th suspected narrow endpoint as an example, the suspected narrow endpoint closest to the h-th suspected narrow endpoint is denoted as the reference suspected narrow endpoint of the h-th suspected narrow endpoint.

[0085] Obtain the distance between the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint of the h-th suspected stenosis endpoint. Calculate the absolute value of the difference between the width of the h-th suspected stenosis endpoint and the width of the reference suspected stenosis endpoint of the h-th suspected stenosis endpoint. Take the inversely proportional normalized value of the product A of the distance and the absolute value of the difference as the probability that the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are the same blood vessel.

[0086] It should be noted that in this embodiment, exp(-A) is used to represent the inverse proportional relationship of A and the normalization process. Implementers can set the inverse proportional function and the normalization function according to the actual situation. exp() is an exponential function with the natural constant as its base. The smaller the difference in blood vessel width between the two endpoints, and the closer the two endpoints are, the greater the probability that they are the same blood vessel.

[0087] Because there are many small endpoints at the tail end of cardiac vessels, some of which have similar widths and are close together, this can interfere with the screening of stenosis areas. However, the directional differences between the two ends of the same vessel are usually small, and the endpoints at the tail end of cardiac vessels diverge outwards. Therefore, this characteristic can be used to calculate the directional consistency between each suspected stenosis endpoint and its reference suspected stenosis endpoint, thereby correcting for the possibility that they belong to the same vessel.

[0088] The preset quantity threshold N is 5, and this will be used as an example for explanation.

[0089] On the skeletal line of the cardiac vascular branch region where any suspected stenosis endpoint is located, starting from the suspected stenosis endpoint, traverse each pixel one by one, and use the least squares method to fit a straight line to the first N pixels traversed to obtain the fitted straight line of the suspected stenosis endpoint.

[0090] The least squares method is a well-known technique, and its specific method will not be introduced here.

[0091] Obtain the minimum angle between the fitted line of the h-th suspected stenosis endpoint and the fitted line of the reference suspected stenosis endpoint of the h-th suspected stenosis endpoint. Use the normalized value of the ratio of the probability that the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are the same vessel to the minimum angle as the final probability that the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are the same vessel.

[0092] It should be noted that in this embodiment, the norm() linear normalization function is used to normalize the ratio of the probability of the two points being the same blood vessel to the minimum angle, normalizing the data value to the interval [0,1]. This is used as an example for explanation. The smaller the minimum angle, the greater the consistency of the extension direction of the skeleton line where the h-th suspected stenosis endpoint and its reference suspected stenosis endpoint are located. Therefore, the probability that the h-th suspected stenosis endpoint and its reference suspected stenosis endpoint are located in the same blood vessel is greater. That is, at this time, the extension directions of the two endpoints are similar, the blood vessel widths corresponding to the two endpoints are similar, and the two endpoints are close to each other.

[0093] The preset matching threshold is 0.8, and we will use this as an example for explanation.

[0094] When the final probability that the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are the same blood vessel is greater than the preset matching threshold, the maximum width between the width of the h-th suspected stenosis endpoint and the width of the reference suspected stenosis endpoint of the h-th suspected stenosis endpoint is obtained. Starting from the straight line segment connecting the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint of the h-th suspected stenosis endpoint, half of the maximum width is extended to both sides from the perpendicular direction of the connecting straight line segment to construct a rectangular region as the second type of cardiovascular stenosis region.

[0095] It should be noted that: when the final probability that the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are the same vessel is less than or equal to a preset matching threshold, it is determined that the h-th suspected stenosis endpoint does not have a corresponding type II cardiovascular stenosis region. Therefore, following the above method, it is determined whether each suspected stenosis endpoint has a corresponding type II cardiovascular stenosis region, resulting in several type II cardiovascular stenosis regions. If the reference suspected stenosis endpoint of suspected stenosis endpoint a is suspected stenosis endpoint b, and suspected stenosis endpoint a has a corresponding type II cardiovascular stenosis region, then when the reference suspected stenosis endpoint of suspected stenosis endpoint b is suspected stenosis endpoint a, then suspected stenosis endpoints a and b have the same corresponding type II cardiovascular stenosis region.

[0096] Step S005: Perform linear transformation on the narrowed area of ​​the heart vessels in the cardiac angiography image to obtain an enhanced cardiac angiography image; perform quantitative analysis on the enhanced cardiac angiography image to obtain the degree of narrowing of the heart vessels.

[0097] Preferably, in one embodiment of the present invention, the method for obtaining the degree of stenosis of the coronary blood vessels includes:

[0098] The first and second types of stenotic areas of the heart and blood vessels are collectively referred to as stenotic areas of the heart and blood vessels.

[0099] In cardiac angiography images, each narrowed area of ​​the coronary artery is linearly transformed to obtain enhanced cardiac angiography images.

[0100] It should be noted that linear transformation is a well-known technique, and the specific method will not be described here. In this embodiment, the intercept of each stenotic region of the heart and blood vessels during linear transformation is 0. The slope of the first type of stenotic region of the heart and blood vessels during linear transformation is the sum of the stenosis probability of the corresponding stenotic skeleton segment plus 1. The slope of the second type of stenotic region of the heart and blood vessels during linear transformation is the sum of the final probability of the corresponding suspected stenosis endpoint and the reference suspected stenosis endpoint being the same vessel plus 1. This is used as an example for description. The intercept and slope are two parameters of the linear transformation, thereby realizing adaptive enhancement of the stenotic region of the heart and blood vessels, improving the clarity of the stenotic region of the heart and blood vessels, and ensuring the accuracy of subsequent vascular stenosis assessment.

[0101] Quantitative angiography is used to process enhanced cardiac angiography images to determine the degree of stenosis in the heart vessels.

[0102] It should be noted that quantitative angiography is a well-known technique for the quantitative analysis of angiographic images, and its specific methods will not be described here. Through precise processing of cardiac angiographic images, quantitative angiography can quantify parameters such as the diameter, length, degree of stenosis, and branching structure of blood vessels. It helps physicians accurately assess vascular lesions, such as atherosclerosis and vascular stenosis, thereby providing data support for clinical decision-making. Quantitative angiography can reduce human error and improve diagnostic accuracy, especially in assessing the severity of vascular lesions and selecting treatment options.

[0103] This invention is now complete.

[0104] In summary, in this embodiment of the invention, a suspected vascular contour region is obtained from a patient's cardiac angiography image. The region is then segmented using branch points on the skeleton lines of this suspected vascular contour region. These segmented skeleton lines further divide the suspected vascular contour region into several suspected vascular branch regions, thereby identifying cardiac vascular branch regions. Based on the changes in vessel width within these branch regions, a first type of cardiac vascular stenosis region is segmented. Based on the similarity between the endpoints of the broken skeleton lines between the cardiac vascular branch regions, a second type of cardiac vascular stenosis region is identified at the break points. A linear transformation is performed on the cardiac vascular stenosis regions in the cardiac angiography image to obtain an enhanced cardiac angiography image. Quantitative analysis of this enhanced image is then performed to obtain the degree of cardiac vascular stenosis. This invention improves the accuracy of vascular stenosis assessment by enhancing the contrast of unclear cardiac vascular stenosis regions.

[0105] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for assessing vascular stenosis based on cardiac angiography images, characterized in that, The method includes the following steps: Obtain cardiac angiography images of the patient; The suspected vascular contour region in the cardiac angiography image is obtained, and the region is segmented by the branch points on the skeleton line of the suspected vascular contour region. The segmented skeleton line is then used to divide the suspected vascular contour region into several suspected vascular branch regions. Based on the shape of the suspected vascular branching region, cardiac vascular branching regions are screened out; based on the changes in the vascular width within the cardiac vascular branching regions, the first type of cardiac vascular stenosis region is segmented. Based on the similarity between the endpoints of the broken skeleton lines between the branch regions of the heart vessels, the second type of heart vessel stenosis region at the break point is identified. Linear transformation is performed on the narrowed areas of the heart vessels in the cardiac angiography images to obtain enhanced cardiac angiography images; quantitative analysis is then performed on the enhanced cardiac angiography images to obtain the degree of narrowing of the heart vessels. Based on the similarity between the endpoints of the broken skeleton lines between cardiac vascular branch regions, the second type of cardiac vascular stenosis region at the break point is determined. The specific steps are as follows: Among the endpoints of the skeleton lines of all suspected vascular contour regions, the endpoints of skeleton lines that are also cardiac vascular branch regions are denoted as suspected stenosis endpoints; the suspected stenosis endpoint closest to the h-th suspected stenosis endpoint is denoted as the reference suspected stenosis endpoint; based on the distance and width between the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint, the probability that the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are the same blood vessel is determined; with a preset threshold N, the skeleton lines of the cardiac vascular branch regions where the suspected stenosis endpoints are located are traversed pixel by pixel starting from the suspected stenosis endpoint, and the least squares method is used to fit a straight line to the first N traversed pixels to obtain the fitted straight line of the suspected stenosis endpoint; based on the angle between the fitted straight lines of the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint and the probability that they are the same blood vessel, the final probability that the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are the same blood vessel is determined; based on the magnitude of the final probability that they are the same blood vessel, the second type of cardiac vascular stenosis region is obtained. Based on the distance and width between the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint, the probability that the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are the same vessel is determined. The specific steps include the following: Calculate the absolute value of the difference between the width of the h-th suspected stenosis endpoint and the width of the reference suspected stenosis endpoint, and take the inverse proportional normalized value of the product of the distance between the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint and the absolute value of the difference as the probability that the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are the same vessel. Based on the angle between the fitted lines of the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint and the probability that they are the same vessel, the final probability that the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are the same vessel is determined. The specific steps include the following: obtaining the minimum angle between the fitted lines of the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint, and taking the normalized value of the ratio of the probability that the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are the same vessel to the minimum angle as the final probability that the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are the same vessel. Linear transformation is performed on narrowed areas of the heart vessels in cardiac angiography images, including: the intercept of the linear transformation is 0 for each narrowed area; the slope of the linear transformation for the first type of narrowed area is the sum of the narrowing probability of the corresponding narrowed skeleton segment plus 1; the slope of the linear transformation for the second type of narrowed area is the sum of the final probability of the corresponding suspected narrowing endpoint and the reference suspected narrowing endpoint being the same vessel plus 1, where the intercept and slope are two parameters of the linear transformation.

2. The method for assessing vascular stenosis based on cardiac angiography images according to claim 1, characterized in that, Based on the shape of suspected vascular branching regions, cardiac vascular branching regions are screened out. The specific steps include the following: The distance between the normal of each pixel on the skeleton line of the suspected blood vessel branch region and the intersection point of the boundary of the suspected blood vessel branch region is used as the width of each pixel on the skeleton line; Based on the width of each pixel on the skeleton line, determine the likelihood of cardiac vessels in suspected vascular branch areas. Based on the likelihood of cardiac blood vessels, the branch regions of cardiac blood vessels are selected.

3. The method for assessing vascular stenosis based on cardiac angiography images according to claim 2, characterized in that, Based on the width of each pixel on the skeleton line, the likelihood of cardiac vessels in suspected vascular branch areas is determined. The specific steps include the following: Obtain the length of the skeleton line of the suspected vascular branch region and the mean of the width of all pixels on the skeleton line. Calculate the ratio of the length to the mean and record it as the first ratio. Then obtain the variance of the gradient values ​​of all pixels on the boundary of the suspected vascular branch region. The normalized value of the ratio of the first ratio to the variance is used as the probability of the cardiac blood vessels in the suspected vascular branch region.

4. The method for assessing vascular stenosis based on cardiac angiography images according to claim 2, characterized in that, Based on the likelihood of identifying cardiac blood vessel branches, the specific steps involved are as follows: Suspected vascular branch regions where the probability of a cardiac vascular branch is greater than a preset vascular threshold are denoted as cardiac vascular branch regions.

5. The method for assessing vascular stenosis based on cardiac angiography images according to claim 2, characterized in that, Based on the changes in vessel width in the branch regions of the heart vessels, the first type of narrowed heart vessels is segmented, including the following specific steps: A preset length threshold L is used to divide the skeleton line of the cardiac blood vessel branch area into several skeleton line segments of length L. The narrowness of a skeleton segment is determined by the difference in the width between pixels on the skeleton segment. Skeleton segments with a narrowing probability greater than a preset narrowing threshold are denoted as narrow skeleton segments; The cardiac vascular branching region is divided into several sub-regions corresponding to skeletal segments. The sub-regions corresponding to the narrow skeletal segments are denoted as the first type of cardiac vascular stenosis region.

6. The method for assessing vascular stenosis based on cardiac angiography images according to claim 5, characterized in that, The possibility of a skeleton line segment being narrow is determined based on the difference in width between pixels on the skeleton line segment. The specific steps include the following: For any skeleton line segment, calculate the average width of all pixels, and use the normalized value of the average of the absolute values ​​of the differences between the width of all pixels and the average width as the narrowness probability of any skeleton line segment.

7. The method for assessing vascular stenosis based on cardiac angiography images according to claim 1, characterized in that, Based on the probability that the stenosis is in the same blood vessel, the second type of coronary artery stenosis is identified, including the following specific steps: When the final probability that the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint are the same blood vessel is greater than the preset matching threshold, the maximum width between the width of the h-th suspected stenosis endpoint and the width of the reference suspected stenosis endpoint is obtained. Starting from the straight line segment connecting the h-th suspected stenosis endpoint and the reference suspected stenosis endpoint, the maximum width is extended to both sides from the perpendicular direction of the connecting straight line segment by half to construct a rectangular region as the second type of cardiac vascular stenosis region.

Images