A method for measuring the geometry of the carotid artery

Abstract

The present invention provides a method for measuring the geometric morphology of the carotid artery, which is used for quantitative analysis of the three-dimensional structure of the carotid artery. This method uses an improved Fast Marching algorithm to automatically generate the parameterized centerlines of each blood vessel branch at the carotid artery bifurcation. The maximum inscribed spheres are defined on the centerlines, and the radius of the inscribed spheres is determined by calculating the minimum distance from each point on the centerline to the blood vessel wall. Based on this, the local blood vessel diameter is calculated, and the Gaussian weighted average method is used to obtain the characteristic diameter of the blood vessel. Based on the parameterized centerlines and the weighted average diameters, a series of carotid artery geometric morphology parameters are calculated, including the weighted blood vessel diameter ratio, the bifurcation angle tensor, and the normalized curvature index. The method of the present invention has a high degree of automation and strong reliability, reduces the human measurement error, and provides an objective and accurate quantitative index for the early diagnosis and risk assessment of carotid artery diseases.

Description

Technical Field

[0001] The present invention belongs to the technical field of medical image processing, and particularly relates to a method for measuring the geometric morphology of the carotid artery. Background Art

[0002] Carotid atherosclerosis lesions are closely related to stroke, so carotid atherosclerosis lesions have always been a hot topic in clinical research; a large number of studies have shown that the anatomy and geometric morphology of blood vessels have a significant impact on hemodynamic characteristics, and may thus become an important risk factor for the early development of atherosclerotic diseases.

[0003] Atherosclerosis is prone to occur at bifurcations, bends, and vascular stenoses in the human arterial system. Due to its special geometric morphology, the carotid bifurcation site has complex and variable blood flow states, becoming a high-incidence area for atherosclerotic plaque formation. Different carotid bifurcation angles will lead to differences in blood flow patterns, especially affecting the size and distribution of the low shear stress area, and low shear stress is considered one of the most dangerous hemodynamic factors in the occurrence and development of arterial diseases, thus affecting the formation and stability of plaques.

[0004] Existing methods for measuring the geometric morphology of the carotid artery have various limitations: First, traditional two-dimensional measurements cannot comprehensively reflect the three-dimensional geometric characteristics of the carotid artery; second, there are differences in the definition of measurement positions in different studies, resulting in incomparable results; third, manual measurement is subjective, increasing measurement errors; more importantly, existing measurement methods often ignore the mutual relationships between geometric parameters such as blood vessel diameter, bifurcation angle, and curvature, and cannot provide a comprehensive assessment of the vascular geometric morphology.

[0005] Therefore, there is an urgent need to develop a three-dimensional reconstruction-based, standardized, and automated quantitative measurement method for the geometric morphology of the carotid artery to provide a more accurate and objective evaluation method and a more reliable basis for carotid atherosclerosis risk assessment and clinical decision-making. Summary of the Invention

[0006] The present invention provides a method for measuring the geometric morphology of the carotid artery. Based on three-dimensional vascular reconstruction and parametric centerline extraction technology, the geometric characteristics of the carotid bifurcation are quantitatively analyzed through an automated algorithm, achieving the purpose of accurately, objectively, and repeatedly evaluating the geometric morphology of the carotid artery and providing a scientific basis for clinical intervention decisions.

[0007] To achieve the above-mentioned invention purpose, the specific technical solutions are as follows:

[0008] A method for measuring the geometric morphology of the carotid artery, the method comprising the following steps:

[0009] Step S1, automatically generate the parameterized centerlines of each blood vessel branch of the three-dimensional carotid artery bifurcation by using the improved Fast Marching method.

[0010] The centerline is represented by the function where is the parameterized length of the blood vessel centerline, The value range of is indicating the total length of the blood vessel centerline to be measured, , , respectively represent the coordinates of the centerline in three-dimensional space.

[0011] Step S2, define the maximum inscribed sphere on each blood vessel centerline, and the radius of the maximum inscribed sphere is determined by calculating the minimum distance from each point on the centerline to the blood vessel wall.

[0012] The radius function , where is a point on the blood vessel wall, represents the distance from each point on the centerline to the blood vessel wall.

[0013] Step S3, calculate the local blood vessel diameter based on the maximum inscribed sphere radius function, and obtain the characteristic diameter of the blood vessel through the weighted average method; the local blood vessel diameter .

[0014] Step S4, calculate the carotid artery geometric shape parameters based on the parameterized centerline and the weighted average diameter, and the geometric shape parameters include the weighted blood vessel diameter ratio, the bifurcation angle tensor, and the normalized curvature index.

[0015] Further, the calculation method of the weighted average diameter in Step S3 is: ; where represents the weighted average diameter of the blood vessel segment to be measured, represents the local blood vessel diameter at the position , is the weight function, and adopts the form of Gaussian weight function: , where represents the center position of the centerline of the blood vessel segment to be calculated, represents the standard deviation of the Gaussian function, controlling the width of the weight distribution, and the value range is to .

[0016] Further, the blood vessel diameter ratio in Step S4 includes the inner total diameter ratio and the outer total diameter ratio ;

[0017] where ; ; , and are the weighted average diameters of the internal carotid artery, external carotid artery, and common carotid artery, respectively;

[0018] The weighted average diameter of the internal carotid artery is measured in the section from 5 mm to 15 mm distal to the center point of the largest inscribed sphere of the internal carotid artery;

[0019] The weighted average diameter of the external carotid artery is measured in the section from 5 mm to 15 mm distal to the center point of the largest inscribed sphere of the external carotid artery;

[0020] The weighted average diameter of the common carotid artery is measured in the section from 5 mm to 15 mm proximal to the center point of the largest inscribed sphere of the common carotid artery.

[0021] Furthermore, in step S4, the bifurcation angle tensor is calculated as follows: ; where is the angle between the tangent vector of the centerline of the internal carotid artery and the tangent vector of the centerline of the common carotid artery in the frontal projection plane, in degrees; represents the angle between the tangent vector of the centerline of the internal carotid artery and the tangent vector of the centerline of the external carotid artery in the frontal projection plane; represents the angle between the tangent vector of the centerline of the internal carotid artery and the tangent vector of the centerline of the common carotid artery in the lateral projection plane; represents the angle between the tangent vector of the centerline of the internal carotid artery and the tangent vector of the centerline of the external carotid artery in the lateral projection plane; the tangent vector is the derivative vector of the centerline parameter length: .

[0022] Furthermore, the calculation of the normalized curvature index in step S4 includes:

[0023] Calculating the local curvature: , representing the Euclidean norm of the second derivative vector of the centerline; calculating the total curvature index: , representing the integral of the local curvature along the entire centerline; calculating the normalized curvature index: ; where is the straight-line distance between two points 1 cm above and below the carotid bifurcation, in millimeters; is the actual length of the centerline between the blood vessels 1 cm above and below the carotid bifurcation, in millimeters; the normalized curvature index is a dimensionless index, and the larger the value, the higher the degree of blood vessel curvature.

[0024] Further, the specific steps of automatically generating the parameterized centerlines of each blood vessel branch of the three-dimensional carotid artery bifurcation by using the improved Fast Marching method in step S1 include:

[0025] Step S11, preprocess the original three-dimensional blood vessel image, and the preprocessing includes noise reduction and enhancement of the blood vessel structure;

[0026] Step S12, perform blood vessel segmentation based on the image gray value to obtain the blood vessel lumen volume;

[0027] Step S13, select a starting point and an ending point in the blood vessel lumen. The starting point is located at the proximal end of the common carotid artery, and the ending points are respectively located at the distal ends of the internal carotid artery and the external carotid artery;

[0028] Step S14, construct a velocity field function based on the blood vessel lumen volume such that the velocity is maximum near the center of the blood vessel and decreases towards the blood vessel wall;

[0029] Step S15, use the velocity field and the Fast Marching method to calculate the shortest time map from the starting point to each point in the blood vessel ;

[0030] Step S16, trace back from the ending point along the gradient direction of the shortest time map to the starting point to obtain the initial path;

[0031] Step S17, smooth the initial path to obtain the parameterized centerline ;

[0032] Step S18, re-parameterize the centerline by the spline interpolation method to ensure that the parameter is proportional to the actual physical length of the centerline.

[0033] Further, the preprocessing of the original three-dimensional blood vessel image in step S11 includes: performing noise reduction on the original image by using the adaptive Gaussian filtering method, and automatically adjusting the filtering kernel parameters according to the local blood vessel diameter; using a blood vessel enhancement algorithm based on the multi-scale Hessian matrix eigenvalue analysis to improve the contrast between the blood vessel and the background; applying the histogram equalization method to adjust the image gray distribution to highlight the blood vessel structure.

[0034] Compared with the prior art, the beneficial effects of the present invention are:

[0035] The present invention measures the blood vessel diameter by introducing the concept of the maximum inscribed sphere and calculates the characteristic diameter using the Gaussian weighted average method, which can more objectively reflect the true size of the blood vessel lumen and reduce the interference of local stenosis or dilation on the measurement results; expands the bifurcation angle from two dimensions to three dimensions, and comprehensively captures the three-dimensional geometric characteristics of the carotid artery bifurcation through the combination description of the front and side angles; comprehensively evaluates the geometric characteristics of the carotid artery through multiple parameters, can more comprehensively reflect its hemodynamic characteristics, and provides a more accurate basis for clinical risk assessment and intervention decision-making; the measurement method of the present invention can be applied to large-sample clinical studies, providing a reliable technical means for exploring the correlation between the geometric morphology of the carotid artery and diseases such as atherosclerosis and stroke. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 is a flowchart of a method for measuring the geometric morphology of the carotid artery according to the present invention; DETAILED DESCRIPTION OF THE INVENTION

[0037] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below. Apparently, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments in the present invention belong to the scope of protection of the present invention.

[0038] It should be noted that it is applicable to quantitative analysis of the geometric morphology of the carotid artery based on three-dimensional medical image data such as computed tomography angiography (CTA), magnetic resonance angiography (MRA) or digital subtraction angiography (DSA), and can be widely used in the clinical diagnosis, risk assessment and academic research of cerebrovascular diseases. The application scenarios include but are not limited to: clinical doctors' assessment of the lesion risk of patients with carotid atherosclerosis, preoperative vascular anatomical assessment for neurosurgery, vascular morphological analysis of stroke patients, and collection and analysis of carotid artery morphological data in related clinical studies. This measurement method can be integrated into the existing PACS system in the hospital or independent medical image processing software to provide a convenient clinical auxiliary diagnosis tool for doctors.

[0039] As Figure 1 shown, it is a flowchart of a method for measuring the geometric morphology of the carotid artery according to the present invention, and the method includes the following steps:

[0040] Step S1, automatically generating a parameterized centerline of each blood vessel branch of the three-dimensional carotid artery bifurcation using an improved Fast Marching method;

[0041] The centerline is represented by the function where is the parameterized length of the blood vessel centerline, The value range of is which represents the total length of the centerline of the blood vessel to be measured, , , and

[0042] respectively represent the coordinates of the centerline in three-dimensional space.

[0043] The centerline of the blood vessel is the skeletal structure that describes the geometric shape of the blood vessel lumen. Accurate centerline extraction is the basis for subsequent blood vessel morphology analysis. The parametric centerline generated by the improved Fast Marching method in this invention has the characteristics of being continuous, smooth, and located at the center of the blood vessel lumen, and can accurately reflect the running path of the blood vessel. In practical applications, for image data with different resolutions, the algorithm can adaptively adjust parameters to ensure the stability and accuracy of centerline extraction. For example, for CTA images with a resolution of 0.5mm×0.5mm×0.5mm, this method can complete the centerline extraction of the carotid artery bifurcation within 5 - 10 seconds, and the average positioning accuracy is better than 0.3mm.

[0043] The specific steps of automatically generating the parametric centerlines of each blood vessel branch of the three-dimensional carotid artery bifurcation by using the improved Fast Marching method include:

[0044] Step S11: Preprocess the original three-dimensional blood vessel image, and the preprocessing includes noise reduction and enhancement of the blood vessel structure;

[0045] Step S12: Segment the blood vessel based on the image gray value to obtain the blood vessel lumen volume;

[0046] Step S13: Select the starting point and the ending point within the blood vessel lumen. The starting point is located at the proximal end of the common carotid artery, and the ending points are respectively located at the distal ends of the internal carotid artery and the external carotid artery;

[0047] Step S14: Construct a velocity field function based on the blood vessel lumen volume such that the velocity is the largest near the center of the blood vessel and decreases towards the blood vessel wall;

[0048] Step S15: Use the velocity field and the Fast Marching method to calculate the shortest time map from the starting point to each point within the blood vessel ;

[0049] Step S16: Trace back from the ending point along the gradient direction of the shortest time map to the starting point to obtain the initial path;

[0050] Step S17: Smooth the initial path to obtain the parametric centerline ;

[0051] Step S18, re-parameterize the centerline by spline interpolation method to ensure that the parameter is proportional to the actual physical length of the centerline.

[0052] Preprocess the original three-dimensional vascular image, including:

[0053] Perform noise reduction on the original image using the adaptive Gaussian filtering method, and the filter kernel parameters are automatically adjusted according to the local vascular diameter; use the vascular enhancement algorithm based on the multi-scale Hessian matrix eigenvalue analysis to improve the contrast between blood vessels and the background; apply the histogram equalization method to adjust the image gray distribution and highlight the vascular structure.

[0054] The Hessian matrix eigenvalue analysis of the present invention is a powerful vascular enhancement technology that can perform selective enhancement according to the geometric characteristics of blood vessels. Tubular structures (such as blood vessels) in three-dimensional images have specific local second-order structural characteristics, that is, the second derivative in the blood vessel direction is close to zero, while the absolute value of the second derivative perpendicular to the blood vessel direction is large and has the same sign. By analyzing the eigenvalues of the Hessian matrix, this feature can be identified, thereby selectively enhancing the vascular structure. In practical applications, for blood vessels of different diameters, Gaussian kernels of different scales are used for convolution calculation of the Hessian matrix, which can enhance the entire range of vascular structures from large blood vessels (such as the common carotid artery, with a diameter of about 6-8 mm) to small blood vessels (such as the distal branches of the external carotid artery, with a diameter of about 2-3 mm), ensuring the integrity of subsequent analysis.

[0055] Step S2, define the maximum inscribed sphere on each blood vessel centerline, and the radius of the maximum inscribed sphere is determined by calculating the minimum distance from each point on the centerline to the blood vessel wall;

[0056] Radius function , where is a point on the blood vessel wall, represents the distance from each point on the centerline to the blood vessel wall;

[0057] Step S3, calculate the local vascular diameter based on the maximum inscribed sphere radius function, and obtain the characteristic diameter of the blood vessel by the weighted average method;

[0058] The maximum inscribed sphere is a method for accurately expressing the local geometric characteristics of blood vessels, which can adapt to the irregularity of the blood vessel cross-sectional shape and is more accurate than the traditional circular cross-section assumption. In practical applications, due to the influence of image resolution and noise, there may be fluctuations in directly calculating the minimum distance from points on the centerline to the blood vessel wall. The present invention uses the multi-directional ray method combined with gradient information for robust distance calculation. Rays are emitted from each point on the centerline in multiple directions (usually 32 - 64 directions) until the blood vessel wall is reached (determined by image gradient or threshold detection). This method has a strong fault tolerance for areas with fuzzy blood vessel wall definition and can significantly improve the accuracy and stability of distance measurement. Clinical verification shows that the consistency between the blood vessel diameters measured by this method and those measured by digital subtraction angiography (DSA) is better than 95%.

[0059] Local blood vessel diameter ; The calculation method of the weighted average diameter is as follows: ; Wherein, represents the weighted average diameter of the blood vessel segment to be measured, represents at the position the local blood vessel diameter, is the weight function, in the form of a Gaussian weight function: ; Wherein, represents the central position of the centerline of the blood vessel segment to be calculated, represents the standard deviation of the Gaussian function, controlling the width of the weight distribution, and the value range is to .

[0060] Step S4, based on the parameterized centerline and the weighted average diameter, calculate the carotid artery geometric shape parameters, and the geometric shape parameters include the weighted blood vessel diameter ratio, the bifurcation angle tensor, and the normalized curvature index.

[0061] The geometric shape parameters are the core indicators for quantitative analysis of the carotid artery morphology and can comprehensively reflect the geometric characteristics of the carotid artery bifurcation and its possible hemodynamic effects. The three types of core parameters (diameter ratio, angle tensor, and curvature index) proposed by the present invention describe the morphological characteristics of the blood vessels from different dimensions and provide multi-dimensional quantitative indicators for subsequent clinical research and risk assessment. The combined analysis of these parameters can reveal the correlation between the carotid artery morphology and diseases. For example, a specific combination of diameter ratio and bifurcation angle may be correlated with the location of plaque formation, while a higher curvature index may be correlated with blood flow disorder and abnormal endothelial cell stress distribution. In a clinical study involving 200 carotid artery CTA images, the correlation analysis between these geometric parameters and plaque formation showed significant statistical significance (p < 0.01).

[0062] The blood vessel diameter ratio includes the internal total diameter ratio Ratio of inner and outer total diameters ; where ; ; , and are the weighted average diameters of the internal carotid artery, external carotid artery, and common carotid artery respectively; the weighted average diameter of the internal carotid artery is measured in the section 5 mm to 15 mm distal from the center point of the largest inscribed sphere of the internal carotid artery; the weighted average diameter of the external carotid artery is measured in the section 5 mm to 15 mm distal from the center point of the largest inscribed sphere of the external carotid artery; the weighted average diameter of the common carotid artery is measured in the section 5 mm to 15 mm proximal from the center point of the largest inscribed sphere of the common carotid artery.

[0063] Bifurcation angle tensor is calculated as: ; where is the angle between the tangent vector of the centerline of the internal carotid artery and the tangent vector of the centerline of the common carotid artery in the frontal projection plane, in degrees; represents the angle between the tangent vector of the centerline of the internal carotid artery and the tangent vector of the centerline of the external carotid artery in the frontal projection plane; represents the angle between the tangent vector of the centerline of the internal carotid artery and the tangent vector of the centerline of the common carotid artery in the lateral projection plane; represents the angle between the tangent vector of the centerline of the internal carotid artery and the tangent vector of the centerline of the external carotid artery in the lateral projection plane; the tangent vector is the derivative vector of the centerline parameter length: .

[0064] Traditional measurement methods often only consider the bifurcation angle in a single plane and cannot capture the three-dimensional characteristics of vascular bifurcations. The bifurcation angle has a significant impact on hemodynamics, especially on the flow field characteristics and wall shear stress distribution at the bifurcation. Computer fluid dynamics (CFD) simulation studies have shown that larger bifurcation angles (especially > 60°) are often accompanied by larger areas of low wall shear stress regions at the origin of the internal carotid artery, and these regions are exactly where plaques are likely to form.

[0065] The bifurcation angle tensor measurement method of the present invention is based on the calculation of the tangent vector of the parametric centerline and has high stability and repeatability; in clinical applications, doctors can quickly identify patients with possible abnormal blood flow patterns through the value of the bifurcation angle tensor and conduct more detailed evaluations and follow-ups.

[0066] The calculation of the normalized curvature index includes: calculating the local curvature: , representing the Euclidean norm of the second derivative vector of the centerline; calculating the total curvature index: , representing the local curvature Integral along the entire centerline; calculating the normalized curvature index: ; where is the straight-line distance between two points 1 cm above and below the carotid bifurcation, in millimeters; is the actual length of the centerline between the blood vessels 1 cm above and below the carotid bifurcation, in millimeters; the normalized curvature index is a dimensionless index, and the larger the value, the higher the degree of blood vessel curvature.

[0067] This index combines the cumulative effect of local curvature (TCI) and the geometric characteristics of overall curvature ( ), comprehensively capturing the characteristics of blood vessel curvature. In clinical studies, it was found that patients with an NCI value greater than 1.4 had a significantly higher risk of plaque formation at the carotid bifurcation than those with an NCI value lower than 1.2. This correlation remained significant after adjusting for traditional risk factors (such as hypertension, diabetes, smoking, etc.). The numerical stability of this measurement method is good, and even in the case of poor image quality, the measurement error is usually controlled within 5%.

[0068] Apply the histogram equalization method to adjust the image gray-scale distribution to make the blood vessel structure more obvious; among them, at the voxel point the Hessian matrix is defined as the second-order partial derivative matrix of the image brightness function :

[0069]

[0070] Perform eigenvalue decomposition on the Hessian matrix to obtain the eigenvalues ( ) and the corresponding eigenvectors ;

[0071] When the conditions are satisfied, and , determine that this point is a tubular structure, where corresponds to the direction vector of the blood vessel;

[0072] Construct a blood vessel enhancement response function based on the eigenvalues:

[0073] ;

[0074] Among them, , , ; are control parameters, respectively used to adjust the sensitivity of blood vessel recognition, the roundness requirement of the blood vessel cross-section, and the degree of suppression of background noise.

[0075] In step S14, the velocity field function is constructed as follows:

[0076] Step S141: First, calculate the distance mapping function from the blood vessel lumen volume to the blood vessel wall , which represents the Euclidean distance from each point in the volume to the nearest blood vessel wall;

[0077] Step S142: Define the velocity field function: ;

[0078] where, The maximum value of the in-lumen distance mapping function represents the radius of the thickest blood vessel segment; is a parameter for controlling the velocity distribution, and its value range is from 1.0 to 5.0;

[0079] Step S143: Introduce an additional constraint term in the blood vessel bifurcation area to make the velocity field smoothly transition in the bifurcation area:

[0080] , where, is a smoothing function in the bifurcation area, which takes the value of 1 in the non-bifurcation area and gradually increases according to the distance from the bifurcation center in the bifurcation area, ensuring that the centerline can smoothly pass through the bifurcation area.

[0081] The velocity field function is a key component in the Fast Marching method, which determines the path characteristics searched by the algorithm. The velocity field function designed in the present invention has a maximum value near the blood vessel center and decreases towards the blood vessel wall, which ensures that the generated path tends to travel along the blood vessel center. The distance mapping function is the basis for constructing the velocity field. It naturally forms a distribution with the blood vessel center as the maximum value by calculating the Euclidean distance from each point to the nearest blood vessel wall.

[0082] The parameter controls the decreasing rate of the velocity from the center to the wall. A larger value will make the velocity field steeper near the center and enhance the attraction of the centerline. In practical applications, for blood vessels with a larger diameter (such as the common carotid artery), taking a value close to 1.0 can generate a smoother velocity field; while for blood vessels with a smaller diameter (such as the distal part of the external carotid artery), taking a value close to 5.0 can enhance the stability of the centerline. It is particularly worth noting the smoothing function , which solves the instability problem of the traditional Fast Marching method when dealing with blood vessel bifurcations, ensures that the centerline can smoothly pass through the bifurcation area, and improves the accuracy of overall geometric shape measurement.

[0083] The carotid artery geometric shape measurement method proposed by the present invention realizes the precise quantitative analysis of the three-dimensional geometric shape of the carotid artery by constructing a complete automated processing flow. This method can not only be applied to clinical diagnosis and risk assessment, but also provide objective and reliable morphological data for the mechanism research of vascular diseases. Through innovative technologies such as parameterized centerline, maximum inscribed sphere, and weighted average diameter, this method solves the limitations of traditional two-dimensional projection measurement and provides a more comprehensive and accurate description of carotid artery morphological characteristics.

[0084] Clinical verification shows that the geometric parameters measured by this method have a significant correlation with the risk of carotid atherosclerosis, providing a new quantitative index for clinical risk stratification and intervention decision-making.

[0085] The specific embodiments described above further elaborate on the purpose, technical solutions, and beneficial effects of the present invention. It should be understood that the above are only specific embodiments of the present invention and are not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for measuring the geometric morphology of the carotid artery, characterized in that: The method comprises the following steps: Step S1, using an improved Fast Marching method to automatically generate parameterized center lines of each vascular branch of the three-dimensional carotid artery bifurcation; The center line is given by the function Indicates that is the parameterized length of the vessel centerline, The value range is , represents the total length of the centerline of the blood vessel to be measured, , , They represent the coordinates of the center line in three-dimensional space respectively; Step S2, defining a maximum inscribed sphere on the centerline of each blood vessel, wherein the radius of the maximum inscribed sphere is determined by calculating the minimum distance from each point on the centerline to the blood vessel wall; Radius function ,in, is a point on the blood vessel wall, Indicates the distance from each point on the centerline to the vessel wall; Step S3, calculating the local blood vessel diameter based on the maximum inscribed sphere radius function, and obtaining the characteristic diameter of the blood vessel by a weighted average method; Local blood vessel diameter ; The weighted average diameter is calculated as: ;in, represents the weighted average diameter of the blood vessel segment to be measured, Indicates at location The local blood vessel diameter at is the weight function, which adopts the Gaussian weight function form: ,in, represents the center position of the center line of the blood vessel segment to be calculated, Represents the standard deviation of the Gaussian function, controls the width of the weight distribution, and has a value range of to ; Step S4, calculating carotid artery geometric parameters based on the parameterized centerline and the weighted average diameter, wherein the geometric parameters include a weighted vessel diameter ratio, a bifurcation angle tensor, and a normalized tortuosity index; Vessel diameter ratio includes internal to total diameter ratio Ratio of outer diameter to ; in, ; ; , and are the weighted mean diameters of the internal carotid artery, external carotid artery, and common carotid artery, respectively; Bifurcation Angle Tensor The calculation formula is: ;in, The angle between the tangent vector of the centerline of the internal carotid artery and the tangent vector of the centerline of the common carotid artery in the frontal projection plane, in degrees; It represents the angle between the tangent vector of the centerline of the internal carotid artery and the tangent vector of the centerline of the external carotid artery in the frontal projection plane; It represents the angle between the tangent vector of the centerline of the internal carotid artery and the tangent vector of the centerline of the common carotid artery in the lateral projection plane; It represents the angle between the tangent vector of the centerline of the internal carotid artery and the tangent vector of the centerline of the external carotid artery in the lateral projection plane; The tangent vector is the derivative vector of the centerline parameter length: ; The calculation of the normalized tortuosity index includes: Calculate the local curvature: , represents the Euclidean norm of the second derivative vector of the centerline; calculate the total curvature index: , which represents the local curvature Integrate along the entire centerline; calculate the normalized tortuosity index: ;in, It is the straight-line distance between two points 1 cm above and below the carotid bifurcation, in millimeters; The actual length of the centerline between the blood vessels 1 cm above and below the carotid bifurcation, in millimeters; Normalized tortuosity index It is a dimensionless index, and the larger the value, the higher the degree of vascular tortuosity.

2. The method for measuring carotid artery geometry according to claim 1, characterized in that: Weighted mean diameter of the internal carotid artery The measurement range is the section from 5mm to 15mm distal to the center point of the largest inscribed sphere of the internal carotid artery; Weighted mean diameter of external carotid artery The measurement range is the section from 5mm to 15mm distal to the center point of the largest inscribed sphere of the external carotid artery; Weighted mean diameter of the common carotid artery The measurement range is the segment from 5mm to 15mm proximal to the center point of the largest internal inscribed sphere of the common carotid artery.

3. The method for measuring carotid artery geometry according to claim 1, characterized in that: The specific steps of automatically generating the parameterized center lines of each vascular branch of the three-dimensional carotid artery bifurcation using the improved Fast Marching method in step S1 include: Step S11, preprocessing the original three-dimensional blood vessel image, wherein the preprocessing includes noise reduction and blood vessel structure enhancement; Step S12, performing blood vessel segmentation based on the image grayscale value to obtain the blood vessel cavity volume; Step S13, selecting a starting point and an end point in the blood vessel cavity, the starting point being located at the proximal end of the common carotid artery, and the end points being located at the distal ends of the internal carotid artery and the external carotid artery respectively; Step S14, constructing a velocity field function based on the vascular cavity volume , so that the velocity is maximum near the center of the blood vessel and decreases toward the blood vessel wall; Step S15, using the velocity field The Fast Marching method is used to calculate the shortest time from the starting point to each point in the blood vessel. ; Step S16, from the end point along the shortest time graph The gradient direction Trace back to the starting point to obtain the initial path; Step S17, smoothing the initial path to obtain a parameterized center line ; Step S18, re-parameterize the center line by spline interpolation method to ensure that the parameters Directly proportional to the actual physical length of the centerline.

4. The method for measuring carotid artery geometry according to claim 3, characterized in that: In step S11, the original three-dimensional blood vessel image is preprocessed, including: The original image was denoised using an adaptive Gaussian filter, and the filter kernel parameters were automatically adjusted according to the local blood vessel diameter. A blood vessel enhancement algorithm based on multi-scale Hessian matrix eigenvalue analysis was used to improve the contrast between the blood vessels and the background. The histogram equalization method was used to adjust the image grayscale distribution and highlight the blood vessel structure.

Images