WO2018068195A1 - Method and device for extracting vessel ridge point on basis of image gradient vector flow field - Google Patents

Abstract

A method and device for extracting a vessel ridge point on the basis of an image gradient vector flow field. The method comprises: enhancing a vessel target in an angiographic image (101); obtaining a gradient vector flow field of the enhanced image by using a gradient vector flow field model (102); and separately calculating cosine values of included angles between a vector at a pixel point of the enhanced image and vectors at two adjacent pixel points of the pixel point along a coordinate axis direction according to the gradient vector flow field of the enhanced image, the pixel point being a vessel ridge point in the coordinate axis direction if the cosine values meet a threshold condition (103). The method can effectively suppress the influence of background noise, improve the extraction efficiency of vessel ridge points in an angiographic image, increase the number of vessel target ridge points to be detected, and improve the detection precision of vessel ridge points, thereby providing a basis for subsequent vascular centerline extraction and vessel modeling.

Description

Blood vessel ridge point extraction method and device based on image gradient vector flow field Technical field

The present application belongs to the field of medical image processing, and in particular, to a method and a device for extracting a vascular ridge point based on an image gradient vector flow field.

Background technique

With the rapid development of medical imaging techniques such as CT angiography and magnetic resonance angiography (MRA), the use of image post-processing techniques to acquire vascular ridge points and centerline information is critical for describing the morphological structure of blood vessels and achieving vascular target reconstruction. Significance, an effective solution to this problem, can improve the automation level and implementation accuracy of vascular disease diagnosis, surgical planning and surgical navigation, thereby improving the efficiency of diagnosis and the success rate of surgery.

其次，根据二阶微分特性：即沿垂直于管状目标径向方向的Hessian矩阵的特征矢量v_i对应的特征值λ_i为负的点，作为存在血管脊点的必要条件。In the prior art, under the standard condition that the tubular target (vessel) is high signal, the background is low signal, and the contour of the tubular target cross section is Gaussian, the local gray maximum point perpendicular to the radial direction of the tubular target is Think of the tubular target ridge point. The traditional method for extracting vascular ridge points is to determine local ridge points by first-order differential and second-order differential of images according to the definition of local gamma maxima. Specifically, first, the extreme point is determined by the first-order differential, and for the pixel in the image, the point where the gradient value is 0 is a sufficient condition of the local extreme point, that is, the satisfaction

Secondly, according to the second-order differential characteristic: the point at which the eigenvalue λ _i corresponding to the eigenvector v _i of the Heessian matrix perpendicular to the radial direction of the tubular target is negative, as a necessary condition for the presence of the vascular ridge point.

The above method for extracting vascular ridge points is very sensitive to the noise existing in the background of the image, and usually erroneously recognizes the local noise bright points as extreme points, thereby greatly increasing the number of non-target ridge points in the image.

Summary of the invention

The present invention provides a method and a device for extracting a vascular ridge point based on an image gradient vector flow field, which is used to solve the problem that the ridge point extraction method based on local extremum point definition is susceptible to high-intensity noise points and is extracted. The problem of low ridge point accuracy is not high.

In order to solve the above technical problem, a technical solution of the present application is to provide a blood vessel ridge point value-based method based on an image gradient vector flow field, including:

Enhancing vascular targets in angiographic images;

The gradient vector flow field model is used to obtain the gradient vector flow field of the enhanced image;

Calculating, according to the gradient vector flow field of the enhanced image, a cosine value of a vector at a pixel point in the enhanced image and a vector angle between two adjacent pixel points of the pixel along a coordinate axis direction, if the cosine value When the threshold condition is satisfied, the pixel point is a vascular ridge point in the direction of the coordinate axis.

Another technical solution of the present application is to provide a blood vessel ridge point extraction device based on an image gradient vector flow field, the device comprising:

An enhancement module for enhancing a blood vessel target in an angiographic image;

a gradient vector flow field calculation module for obtaining a gradient vector flow field of the enhanced image by using a gradient vector flow field model;

The ridge point detecting module calculates, according to the gradient vector flow field of the enhanced image, a cosine value of a vector at a pixel point in the enhanced image and a vector angle between two adjacent pixel points of the pixel along a coordinate axis direction, If the cosine value satisfies a threshold condition, the pixel point is a vascular ridge point in the direction of the coordinate axis.

The application can effectively suppress the influence of background noise, improve the extraction efficiency of vascular ridge points in angiographic images, increase the number of detection of vascular target ridge points and improve the detection accuracy of vascular ridge points, and provide a basis for subsequent extraction of blood vessel centerline and blood vessel modeling. .

DRAWINGS

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings used in the description of the embodiments will be briefly described below. It is obvious that the drawings in the following description are only some embodiments of the present application, Those skilled in the art can also obtain other drawings based on these drawings without paying any creative work.

1 is a flowchart of a method for extracting a vascular ridge point based on an image gradient vector flow field according to an embodiment of the present application;

2a is a schematic view of an isolated ridge point according to an embodiment of the present application;

2b is a schematic view of an isolated ridge group in the embodiment of the present application;

3a is a magnetic resonance three-dimensional angiography image of an embodiment of the present application;

Figure 3b is a maximum density projection image of the image data processed by the multi-scale vascular enhancement function of Figure 3a;

4a is an image of a blood vessel enhanced embodiment of the present application;

4b is a schematic diagram of a gradient vector flow field for obtaining an image using the gradient vector flow field model for the image shown in FIG. 4a;

FIG. 5a is a three-dimensional simulation diagram of an angiographic image according to an embodiment of the present application; FIG.

Figure 5b is a result of the ridge point extraction of the image shown in Figure 5a;

FIG. 6 is a structural diagram of a blood vessel ridge point extracting apparatus based on an image gradient vector flow field according to an embodiment of the present application.

detailed description

In order to make the technical features and effects of the present application more obvious, the technical solutions of the present application are further described below with reference to the accompanying drawings, and the present application may also be described or implemented in various other specific examples, and any person skilled in the art is in the scope of the claims. Equivalent transformations made within the scope of protection of this application.

As shown in FIG. 1 , FIG. 1 is a flowchart of a method for extracting a vascular ridge point based on an image gradient vector flow field according to an embodiment of the present application. The embodiment can effectively suppress the influence of background noise, improve the extraction efficiency of vascular ridge points in angiographic images, increase the number of detection of vascular target ridge points, and improve the detection accuracy of vascular ridge points, and provide for subsequent extraction and vascular modeling of blood vessel centerlines. basis.

Specifically, the method includes:

Step 101: Enhance the blood vessel target in the angiographic image.

Step 102: Calculate a gradient vector flow field of the enhanced image by using a gradient vector flow field model.

The gradient vector flow field model is a globally optimized vector field. The vector of each point (pixel point) in the gradient vector flow field of the image points to the ridge point of the blood vessel target, that is, the local gray maximum point of the radial direction of the blood vessel target. According to this feature, the ridge points in the angiographic image can be accurately, stably and quickly extracted by the following step 103.

Step 103: Calculate, according to the gradient vector flow field of the enhanced image, respectively, a cosine value of a vector at a pixel point in the enhanced image and a vector angle between two adjacent pixel points of the pixel along a coordinate axis direction. When the cosine value satisfies the threshold condition, the pixel point is a vascular ridge point in the direction of the coordinate axis.

The angiographic image described in the present application may be a two-dimensional image or a three-dimensional image. For a two-dimensional angiographic image, the coordinate axis direction includes a ±x, ±y direction, and for a three-dimensional angiographic image, the coordinate axis direction includes ±x, ±y, ±z directions. In practice, step 103 is repeated until all pixel points of the enhanced image and all coordinate axes are traversed to obtain all vascular ridge points.

In one embodiment, after obtaining all of the vascular ridge points, the method further comprises: removing the isolated points from the obtained vascular ridge points. In detail, the process of removing isolated points from all obtained vascular ridge points includes:

Judging whether a certain ridge point A has other ridge points in the circle formed by two concentric circles whose center is the radius d and d+d _0. If not, the ridge point A is the center of the circle The ridge point in the circle formed by the radius d is deleted, wherein d and d ₀ are distance constants.

As shown in Fig. 2a and Fig. 2b, in Fig. 2a, if there are no other ridge points in the ring, the isolated ridge point A is deleted; in Fig. 2b, if there are no other ridge points in the ring, the ridge points A, B are The isolated ridge group consisting of C and D is deleted. In practice, the values of the distance constants d and d ₀ can be determined according to the visual subjective grasp of isolated points, combined with the correlation of noise pseudo ridge point distribution, real ridge point distribution and ridge line characteristics, generally taking d=5~ 8 pixel distances, d ₀ = 3 to 4 pixel distances.

In one embodiment, for a three-dimensional angiographic image, the above step 101 may enhance the vascular target in the three-dimensional angiographic image by the first multi-scale vascular enhancement function as follows:

R_A、R_B和S为三个测度函数；R_A用来区分片状和线状结构；R_B用来区分点状结构和线状结构；S用于区分背景像素；α、β和c为阈值，用于控制血管增强算法对R_A、R_B和S的敏感性；λ₁、λ₂和λ₃为Hessian矩阵H的三个特征值，且满足|λ₁|≤|λ₂|≤|λ₃|，D为图像的维度；s为某个像素点。v ₀ (s) is a first multi-scale vascular enhancement function;

R _A , R _B and S are three measure functions; R _{A is} used to distinguish between sheet and line structures; R _{B is} used to distinguish between point structures and line structures; S is used to distinguish background pixels; α, β and c The threshold is used to control the sensitivity of the vascular enhancement algorithm to R _A , R _B and S; λ ₁ , λ ₂ and λ ₃ are the three eigenvalues of the Hessian matrix H and satisfy |λ ₁ | ≤ | λ ₂ | ≤|λ ₃ |, D is the dimension of the image; s is a certain pixel point.

It should be noted that the Hessian matrix H is a square matrix composed of a third-order partial derivative, which can be calculated by an existing method. The thresholds α and β are generally taken as 0.5, and the value of the threshold c depends on the gray scale range of the image. Usually takes half of the maximum Hessian matrix norm.

For a two-dimensional angiographic image, step 101 above enhances the vascular target in the two-dimensional angiographic image by a second multi-scale vascular enhancement function as follows:

R_B'和S'为测度函数，R_B'用来区分点状结构和线状结构，S'用于区分背景像素，β和c为阈值，λ₁和λ₂为Hessian矩阵H的二个特征值，且满足|λ₁|≤|λ₂|，D为图像的维度。Where v ₀ '(s) is a second multi-scale vascular enhancement function,

R _B ' and S' are measure functions, R _B ' is used to distinguish between point structures and line structures, S' is used to distinguish background pixels, β and c are threshold values, and λ ₁ and λ _{2 are two} of Hessian matrix H The feature value, and satisfies |λ ₁ | ≤ | λ ₂ |, D is the dimension of the image.

After the angiographic image is processed by the multi-scale vascular enhancement function of equation (1) or (2), the vascular target gray value is enhanced and the background noise is suppressed.

The multi-scale vascular enhancement function is obtained by multi-scale Gaussian filtering and Hessian matrix eigenvalue calculation of angiographic images. The specific calculation process refers to Frangi, A.F., Niessen, W.J., Vincken, K.L., Viergever, M.A., 1998. Multiscale vessel enhancement filtering. In: Proceedings of the International Conference on Medical Image Computing Computer Assisted Intervention. Lect. Notes Comp. Sci., 1496, pp. 130-137, which is not described in detail herein.

In one embodiment, as shown in FIGS. 3a and 3b, FIG. 3a is a magnetic resonance three-dimensional angiography image of the embodiment of the present application, and FIG. 3b is a maximum density projection image of the image data processed by the blood vessel enhancement function of FIG. 3a. As can be seen from Fig. 3a and Fig. 3b, the background noise of the image processed by the multi-scale vascular enhancement function is suppressed, and the point on the center line of the blood vessel has the maximum gray value perpendicular to the direction of the blood vessel.

In detail, the gradient vector flow field is defined as V(x, y) = (u(x, y), v(x, y)), where u(x, y), v(x, y) are respectively The two components of the vector V can be obtained by minimizing the energy functional. The specific formula is as follows:

为f(x,y)的梯度，μ为控制参数。Where ε is the energy functional; (x, y) is the coordinate of the pixel of the enhanced image; u _x , u _y , v _x , v _y are the first-order partial derivatives of the components u and v respectively for x and y; (x, y) is the edge of the enhanced image;

For the gradient of f(x, y), μ is the control parameter.

μ can be set according to the quality of the image (such as noise). Generally, the smaller the μ value, the smaller the dynamic range of the gradient vector flow field, which can detect finer blood vessels, but the more susceptible it is to noise.

Using the variational method, the gradient vector flow field satisfies the following Euler equations:

为laplacian算子，μ为控制参数。Where f _x (x, y) and f _y (x, y) are the values of the blood vessel edge function f(x, y) in the x and y directions of the enhanced image, u(x, y), v(x, y) is the two components of the gradient vector flow field V(x, y) = (u(x, y), v(x, y)), (x, y) is the coordinates of the pixel points of the enhanced image,

For the laplacian operator, μ is the control parameter.

For the two-dimensional angiography image, in the above step 102, using the gradient vector flow field model to obtain the gradient vector flow field of the enhanced image further includes:

Solving the above formula (4), the iterative formula of the two components of the gradient vector flow field of the enhanced two-dimensional angiography image is:

Where n is the number of iterations.

Similarly, the iterative formula for the three components of the gradient vector flow field of the enhanced three-dimensional angiography image is:

The convergence value of the formula (5) is obtained, and the gradient vector flow field of the enhanced two-dimensional angiography image is obtained:

V(x, y) = (u(x, y), v(x, y)).

The convergence value of the formula (6) is obtained, and the gradient vector flow field of the enhanced three-dimensional angiography image is obtained:

V(x, y, z) = (u(x, y, z), v(x, y, z), w(x, y, z)).

In a specific embodiment, as shown in FIGS. 4a and 4b, FIG. 4a is a blood vessel enhanced image of the embodiment of the present application, and FIG. 4b is a gradient vector flow for obtaining an image using the gradient vector flow field model for the image shown in FIG. 4a. A schematic diagram of the field, as can be seen from Figure 4b, the vector arrow points to the vascular target centerline, the gradient at the centerline of the vessel target is strongest, and the gradient field away from the centerline is zero.

The process of detecting the ridge point in the above step 103 will be described below in two specific embodiments:

(1) Two-dimensional angiography image ridge detection

For the x-axis direction of the two-dimensional angiography image, it is determined whether a certain pixel point satisfies the following formula, and if satisfied, the pixel point is a ridge point in the x-axis direction:

Where (i, j) is the coordinates of a certain pixel point, (i-1, j) and (i+1, j) are (i, j) two adjacent pixel points in the x-axis direction, V _x ^{( i, j)} , V _x ^{(i-1, j)} , V _x ^{(i+1, j)} are (i, j), (i-1, j), (i+1, j) pixel points, respectively The components of the vector in the x-axis direction, V _y ^(i,j) , V _y ^(i-1,j) , V _y ^(i+1,j) are (i,j), (i-1,j), respectively. The component of the (i+1,j) pixel at the y-axis direction, and T ₀ is the threshold.

For the y-axis direction of the two-dimensional angiography image, it is determined whether a certain pixel point satisfies the following formula, and if satisfied, the pixel point is a ridge point in the y-axis direction:

Where (i, j) is the coordinates of a certain pixel point, (i, j-1) and (i, j+1) are (i, j) two adjacent pixel points in the y-axis direction, V _x ^{( i, j)} , V _x ^{(i, j-1)} , V _x ^{(i, j+1)} are (i, j), (i, j-1), (i, j+1) pixel points, respectively The component of the vector in the x-axis direction, V _y ^(i,j) , V _y ^(i-1,j) , V _y ^(i,j+1) are (i,j), (i,j-1) The component of the (i, j+1) pixel at the y-axis direction, and T ₀ is the threshold.

It should be noted that the vector at each pixel point is a normalized vector, and the modulus of the vector is 1. The threshold can be selected according to the extraction precision, and usually T ₀ selects a value in the range [ ₀ , 0.5].

(2) Three-dimensional angiography image ridge detection

For the x-axis direction of the three-dimensional angiography image, it is determined whether a certain pixel point satisfies the following formula, and if satisfied, the pixel point is a ridge point in the x-axis direction:

Where (i, j, k) is the coordinates of a certain pixel point; (i-1, j, k) and (i+1, j, k) are two (i, j, k) in the x-axis direction Neighboring pixels; V _x ^{(i, j, k)} , V _x ^{(i-1, j, k)} , V _x ^{(i+1, j, k)} are (i, j, k), (i -1, j, k), (i+1, j, k) components of the vector in the x-axis direction; V _y ^{(i, j, k)} , V _y ^{(i-1, j, k)} , V _y ^{(i+1, j, k)} are the components of the vector at the (i, j, k), (i-1, j, k), (i+1, j, k) pixel points in the y-axis direction, respectively. ;V _z ^(i,j,k) , V _z ^(i-1,j,k) , V _z ^(i+1,j,k) are (i,j,k), (i-1,j, respectively , k), (i+1, j, k) The component of the vector at the pixel point in the z-axis direction; T ₀ is the threshold.

For the y-axis direction of the three-dimensional angiography image, it is determined whether a certain pixel point satisfies the following formula, and if satisfied, the pixel point is a ridge point in the y-axis direction:

Where (i, j, k) is the coordinates of a certain pixel point; (i, j-1, k) and (i, j+1, k) are two (i, j, k) in the y-axis direction Neighboring pixels; V _x ^{(i, j, k)} , V _x ^{(i, j-1, k)} , V _x ^{(i, j+1, k)} are (i, j, k), (i , j-1, k), (i, j+1, k) the component of the vector in the x-axis direction; V _y ^{(i, j, k)} , V _y ^{(i, j-1, k)} , V _y ^{(i, j+1, k)} are the components of the vector at the (i, j, k), (i, j-1, k), (i, j+1, k) pixel points in the y-axis direction, respectively. ; V _z ^{(i, j, k)} , V _z ^{(i, j-1, k)} , V _z ^{(i, j+1, k)} are (i, j, k), (i, j-1, respectively) , k), (i, j+1, k) the component of the vector in the z-axis direction at the pixel point; T ₀ is the threshold.

For the z-axis direction of the three-dimensional angiography image, it is determined whether a certain pixel point satisfies the following formula, and if satisfied, the pixel point is a ridge point in the z-axis direction:

Where (i, j, k) is the coordinates of a certain pixel point; (i, j, k-1) and (i, j, k+1) are two (i, j, k) in the z direction Neighboring pixels; V _x ^{(i, j, k)} , V _x ^{(i, j, k-1)} , V _x ^{(i, j, k+1)} are (i, j, k), (i, respectively j, k-1), (i, j, k+1) the component of the vector at the x-axis direction; V _y ^{(i, j, k)} , V _y ^{(i, j, k-1)} , V _y ^{(i, j, k+1)} are components of the vector at the (i, j, k), (i, j, k-1), (i, j, k+1) pixel points in the y-axis direction, respectively; V _z ^(i,j,k) , V _z ^(i,j,k-1) , V _z ^(i,j,k+1) are (i,j,k), (i,j,k-, respectively) 1), the component of the vector at the (i, j, k+1) pixel point in the z-axis direction; T ₀ is a threshold value.

Similarly, the vector at each pixel is a normalized vector whose modulus is one. The threshold can be selected according to the extraction precision, and usually T ₀ selects a value in the range [ ₀ , 0.5].

In a specific embodiment, as shown in FIG. 5a and FIG. 5b, the ridge point extraction method in the three-dimensional simulation diagram shown in FIG. 5a is extracted by using the image gradient vector flow field based vascular ridge point extraction method described in the present application, and the ridge point extraction result is obtained. As shown in Figure 5b.

Based on the same inventive concept, an embodiment of the present application further provides a blood vessel ridge extraction device based on an image gradient vector flow field, as described in the following embodiments. Since the principle of solving the problem of the device is similar to the method for extracting the vascular ridge point, the implementation of the device can be referred to the implementation of the method for extracting the vascular ridge point, and the repetition will not be repeated.

As shown in FIG. 6, FIG. 6 is a vascular ridge point extracting device based on an image gradient vector flow field according to an embodiment of the present application. The device can be implemented in a smart terminal, such as a mobile phone, a tablet computer, or the like by a logic circuit, or can implement functions of various components by software in a functional module manner, and run on the smart terminal. Specifically, the device includes: an enhancement module 601, configured to enhance a blood vessel target in the angiographic image;

a gradient vector flow field calculation module 602, configured to obtain a gradient vector flow field of the enhanced image by using the gradient vector flow field model;

The ridge point detecting module 603 respectively calculates a cosine value of a vector at a pixel point in the enhanced image and a vector angle between two adjacent pixel points of the pixel along a coordinate axis according to the gradient vector flow field of the enhanced image. If the cosine value satisfies a threshold condition, the pixel point is a vascular ridge point in the direction of the coordinate axis.

Further, the ridge point extracting device further includes: a culling module 604, configured to remove the isolated points from all the obtained vascular ridge points.

When implemented, the culling module 604 is specifically configured to determine whether a ridge point has other ridge points in a ring circle composed of two concentric circles having a radius d and d+d ₀ as a center, if not If there is, the ridge point is the center of the circle, and the ridge point of the circle range formed by the radius d is deleted, wherein d and d ₀ are distance constants.

In one embodiment, the enhancement module 601 enhances blood vessels in a three-dimensional angiographic image by a first multi-scale vascular enhancement function as follows:

R_A、R_B和S为测度函数，R_A用来区分片状和线状结构，R_B用来区分点状结构和线状结构，S用于区分背景像素，α、β和c为阈值，λ₁、λ₂和λ₃为Hessian矩阵H的三个特征值，且满足|λ₁|≤|λ₂|≤|λ₃|，D为图像的维度；v ₀ (s) is the first multi-scale vascular enhancement function,

R _A , R _B and S are measure functions, R _{A is} used to distinguish between sheet and line structures, R _{B is} used to distinguish between point structures and line structures, S is used to distinguish background pixels, and α, β and c are threshold values. , λ ₁ , λ ₂ and λ ₃ are three eigenvalues of the Hessian matrix H, and satisfy |λ ₁ | ≤ | λ ₂ | ≤ | λ ₃ |, D is the dimension of the image;

The vascular target in the two-dimensional angiography image is enhanced by a second multi-scale vascular enhancement function as follows:

R_B'和S'为测度函数，R_B'用来区分点状结构和线状结构，S'用于区分背景像素，β和c为阈值，λ₁和λ₂为Hessian矩阵H的二个特征值，且满足|λ₁|≤|λ₂|，D为图像的维度。Where v ₀ '(s) is a second multi-scale vascular enhancement function,

R _B ' and S' are measure functions, R _B ' is used to distinguish between point structures and line structures, S' is used to distinguish background pixels, β and c are threshold values, and λ ₁ and λ _{2 are two} of Hessian matrix H The feature value, and satisfies |λ ₁ | ≤ | λ ₂ |, D is the dimension of the image.

In an embodiment, the gradient vector flow field calculation module 602 is specifically configured to:

Solve the Euler equations of the gradient vector flow field model as follows:

The iterative formula of the two components of the gradient vector flow field with respect to time after the enhanced two-dimensional angiography image is:

为laplacian算子，μ为控制参数； Where n is the number of iterations, f _x (x, y) and f _y (x, y) are the values of the vessel edge function f(x, y) in the x and y directions of the enhanced two-dimensional angiography image, u( x, y), v(x, y) are the two components of the gradient vector flow field V(x, y) = (u(x, y), v(x, y)), and (x, y) is enhanced The coordinates of the pixel points of the posterior two-dimensional angiography image,

For the laplacian operator, μ is the control parameter;

Similarly, the iterative formula for the three components of the gradient vector flow field of the enhanced three-dimensional angiography image is:

为laplacian算子，μ为控制参数。Where n is the number of iterations, f _x (x, y, z), f _y (x, y, z) and f _z (x, y, z) are the vascular edge functions of the enhanced three-dimensional angiography image f(x , y, z) values in the x, y, and z directions, (x, y, z) are the coordinates of the pixel points of the enhanced three-dimensional angiography image,

For the laplacian operator, μ is the control parameter.

For the ridge point detection module 603 to implement the ridge point detection process, refer to the above method embodiment, which is not described herein again.

The technical solution described in the above embodiments of the present application can effectively suppress the influence of background noise, improve the extraction efficiency of vascular ridge points in angiographic images, increase the number of detection of vascular target ridge points, and improve the detection accuracy of vascular ridge points, which is a follow-up vascular center. Line extraction and vascular modeling provide the basis.

The embodiment of the present application further provides an electronic device, including a processor and a memory including a computer readable program, when executed, causing the processor to execute the image gradient vector based flow described in the above embodiment Field ridge point extraction method.

The embodiment of the present application further provides a computer readable program, wherein when the program is executed in an electronic device, the program causes a computer to execute an image gradient vector based flow field in the electronic device as described in the above embodiment Ridge extraction method.

The embodiment of the present application further provides a storage medium storing a computer readable program, wherein the computer readable program causes the computer to execute the image gradient vector flow field based ridge point extraction method described in the above embodiments in the electronic device.

It should be understood that portions of the application can be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or combination of the following techniques well known in the art: having logic gates for implementing logic functions on data signals A discrete logic circuit of a circuit, an application specific integrated circuit with a suitable combination of logic gates, a programmable gate array (PGA), a field programmable gate array (FPGA), and the like.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (system), and computer program products according to embodiments of the present application. It will be understood that each flow and/or block of the flowchart illustrations and/or FIG. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing device to produce a machine for the execution of instructions for execution by a processor of a computer or other programmable data processing device. Means for implementing the functions specified in one or more of the flow or in a block or blocks of the flow chart.

The computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture comprising the instruction device. The apparatus implements the functions specified in one or more blocks of a flow or a flow and/or block diagram of the flowchart.

These computer program instructions can also be loaded onto a computer or other programmable data processing device such that a series of operational steps are performed on a computer or other programmable device to produce computer-implemented processing for execution on a computer or other programmable device. The instructions provide steps for implementing the functions specified in one or more of the flow or in a block or blocks of a flow diagram.

The above description is only for explaining the technical solutions of the present application, and those skilled in the art can modify and change the above embodiments without departing from the spirit and scope of the present application. Therefore, the scope of protection of the application should be determined by the scope of the claims.

Claims (14)

A method for extracting a vascular ridge point based on an image gradient vector flow field, comprising: Enhancing vascular targets in angiographic images; The gradient vector flow field model is used to obtain the gradient vector flow field of the enhanced image; Calculating, according to the gradient vector flow field of the enhanced image, a cosine value of a vector at a pixel point in the enhanced image and a vector angle between two adjacent pixel points of the pixel along a coordinate axis direction, if the cosine value When the threshold condition is satisfied, the pixel point is a vascular ridge point in the direction of the coordinate axis. The vascular ridge point extraction method according to claim 1, wherein after obtaining all the vascular ridge points, the method further comprises: removing the isolated points from all the obtained vascular ridge points. The vascular ridge point extraction method according to claim 2, wherein the process of culling the isolated points from the obtained vascular ridge points further comprises: Judging whether a ridge point has other ridge points in the circle formed by two concentric circles whose center is the radius d and d+d _0. If not, the ridge point is the center and radius The ridge point deletion within the circle formed for d, where d, d ₀ are distance constants. The vascular ridge point extraction method according to claim 1, wherein the vascular target in the three-dimensional angiographic image is enhanced by the first multi-scale vascular enhancement function as follows:

v ₀ (s) is the first multi-scale vascular enhancement function,

R _A , R _B and S are measure functions, R _{A is} used to distinguish between sheet and line structures, R _{B is} used to distinguish between point structures and line structures, S is used to distinguish background pixels, and α, β and c are threshold values. , λ ₁ , λ ₂ and λ ₃ are three eigenvalues of the Hessian matrix H, and satisfy |λ ₁ | ≤ | λ ₂ | ≤ | λ ₃ |, D is the dimension of the image; The vascular target in the two-dimensional angiography image is enhanced by a second multi-scale vascular enhancement function as follows:

Where v ₀ '(s) is a second multi-scale vascular enhancement function,

RB' and S' are the measure functions, R _B ' is used to distinguish the point structure and the line structure, S' is used to distinguish the background pixels, β and c are the threshold values, and λ ₁ and λ ₂ are the two characteristics of the Hessian matrix H. Value, and satisfy |λ ₁ | ≤ | λ ₂ |, D is the dimension of the image. The vascular ridge point extraction method according to claim 1, wherein the gradient vector flow field model is used to obtain the gradient vector flow field of the enhanced image further comprising: Solve the Euler equations of the gradient vector flow field model as follows:

The iterative formula for the two components of the gradient vector flow field of the enhanced two-dimensional angiography image is:

Where n is the number of iterations, f _x (x, y) and f _y (x, y) are the values of the vessel edge function f(x, y) in the x and y directions of the enhanced two-dimensional angiography image, u( x, y), v(x, y) are the two components of the gradient vector flow field V(x, y) = (u(x, y), v(x, y)), and (x, y) is enhanced The coordinates of the pixel points of the posterior two-dimensional angiography image,

For the laplacian operator, μ is the control parameter; Similarly, the iterative formula for the three components of the gradient vector flow field of the enhanced three-dimensional angiography image is:

Where n is the number of iterations, f _x (x, y, z), f _y (x, y, z) and f _z (x, y, z) are the vascular edge functions of the enhanced three-dimensional angiography image f(x , y, z) values in the x, y, and z directions, (x, y, z) are the coordinates of the pixel points of the enhanced three-dimensional angiography image,

For the laplacian operator, μ is the control parameter. The vascular ridge point extraction method according to claim 1, wherein the vector at a pixel point in the enhanced image and the two pixels along the direction of a coordinate axis are respectively calculated according to the gradient vector flow field of the enhanced image. Neighbor a cosine value of a vector angle at a pixel point. If the cosine value satisfies a threshold condition, the vascular ridge point of the pixel point in the direction of the coordinate axis further includes: For the x-axis direction of the two-dimensional angiography image, it is determined whether a certain pixel point satisfies the following formula, and if satisfied, the pixel point is a ridge point in the x-axis direction:

Where (i, j) is the coordinates of a certain pixel point, (i-1, j) and (i+1, j) are (i, j) two adjacent pixel points in the x-axis direction, V _x ^{( i, j)} , V _x ^{(i-1, j)} , V _x ^{(i+1, j)} are (i, j), (i-1, j), (i+1, j) pixel points, respectively The components of the vector in the x-axis direction, V _y ^(i,j) , V _y ^(i-1,j) , V _y ^(i+1,j) are (i,j), (i-1,j), respectively. a component of the (i+1,j) pixel at the y-axis direction, and T ₀ is a threshold; For the y-axis direction of the two-dimensional angiography image, it is determined whether a certain pixel point satisfies the following formula, and if satisfied, the pixel point is a ridge point in the y-axis direction:

Where (i, j) is the coordinates of a certain pixel point, (i, j-1) and (i, j+1) are (i, j) two adjacent pixel points in the y-axis direction, V _x ^{( i, j)} , V _x ^{(i, j-1)} , V _x ^{(i, j+1)} are (i, j), (i, j-1), (i, j+1) pixel points, respectively The component of the vector in the x-axis direction, V _y ^(i,j) , V _y ^(i-1,j) , V _y ^(i,j+1) are (i,j), (i,j-1) The component of the (i, j+1) pixel at the y-axis direction, and T ₀ is the threshold. The vascular ridge point extraction method according to claim 1, wherein the vector at a pixel point in the enhanced image and the two pixels along the direction of a coordinate axis are respectively calculated according to the gradient vector flow field of the enhanced image. The cosine of the angle of the vector at the adjacent pixel, if the cosine value satisfies the threshold condition, the vascular ridge point of the pixel in the direction of the coordinate axis further includes: For the x-axis direction of the three-dimensional angiography image, it is determined whether a certain pixel point satisfies the following formula, and if satisfied, the pixel point is a ridge point in the x-axis direction:

Where (i, j, k) is the coordinates of a certain pixel point; (i-1, j, k) and (i+1, j, k) are two (i, j, k) in the x-axis direction Neighboring pixels; V _x ^{(i, j, k)} , V _x ^{(i-1, j, k)} , V _x ^{(i+1, j, k)} are (i, j, k), (i -1, j, k), (i+1, j, k) components of the vector in the x-axis direction; V _y ^{(i, j, k)} , V _y ^{(i-1, j, k)} , V _y ^{(i+1, j, k)} are the components of the vector at the (i, j, k), (i-1, j, k), (i+1, j, k) pixel points in the y-axis direction, respectively. ;V _z ^(i,j,k) , V _z ^(i-1,j,k) , V _z ^(i+1,j,k) are (i,j,k), (i-1,j, respectively , k), (i+1, j, k) a component of the vector at the pixel point in the z-axis direction; T ₀ is a threshold; For the y-axis direction of the three-dimensional angiography image, it is determined whether a certain pixel point satisfies the following formula, and if satisfied, the pixel point is a ridge point in the y-axis direction:

Where (i, j, k) is the coordinates of a certain pixel point; (i, j-1, k) and (i, j+1, k) are two (i, j, k) in the y-axis direction Neighboring pixels; V _x ^{(i, j, k)} , V _x ^{(i, j-1, k)} , V _x ^{(i, j+1, k)} are (i, j, k), (i , j-1, k), (i, j+1, k) the component of the vector in the x-axis direction; V _y ^{(i, j, k)} , V _y ^{(i, j-1, k)} , V _y ^{(i, j+1, k)} are the components of the vector at the (i, j, k), (i, j-1, k), (i, j+1, k) pixel points in the y-axis direction, respectively. ; V _z ^{(i, j, k)} , V _z ^{(i, j-1, k)} , V _z ^{(i, j+1, k)} are (i, j, k), (i, j-1, respectively) , k), (i, j+1, k) a component of the vector at the pixel point in the z-axis direction; T ₀ is a threshold; For the z-axis direction of the three-dimensional angiography image, it is determined whether a certain pixel point satisfies the following formula, and if satisfied, the pixel point is a ridge point in the z-axis direction:

Where (i, j, k) is the coordinates of a certain pixel point; (i, j, k-1) and (i, j, k+1) are two (i, j, k) in the z direction Neighboring pixels; V _x ^{(i, j, k)} , V _x ^{(i, j, k-1)} , V _x ^{(i, j, k+1)} are (i, j, k), (i, respectively j, k-1), (i, j, k+1) the component of the vector at the x-axis direction; V _y ^{(i, j, k)} , V _y ^{(i, j, k-1)} , V _y ^{(i, j, k+1)} are components of the vector at the (i, j, k), (i, j, k-1), (i, j, k+1) pixel points in the y-axis direction, respectively; V _z ^(i,j,k) , V _z ^(i,j,k-1) , V _z ^(i,j,k+1) are (i,j,k), (i,j,k-, respectively) 1), the component of the vector at the (i, j, k+1) pixel point in the z-axis direction; T ₀ is a threshold value. A vascular ridge point extraction device based on an image gradient vector flow field, comprising: An enhancement module for enhancing a blood vessel target in an angiographic image; a gradient vector flow field calculation module for obtaining a gradient vector flow field of the enhanced image by using a gradient vector flow field model; a ridge point detecting module, configured to calculate, according to the gradient vector flow field of the enhanced image, a cosine of a vector at a pixel point in the enhanced image and a vector angle between two adjacent pixel points of the pixel along a coordinate axis direction a value, if the cosine value satisfies a threshold condition, the pixel point is a vascular ridge point in the direction of the coordinate axis. A vascular ridge extraction apparatus according to claim 8, further comprising: a culling module for culling isolated points from all of the obtained vascular ridge points. The vascular ridge extraction device according to claim 9, wherein the culling module is specifically configured to determine that a ridge point is composed of two concentric circles having a center of a circle and a radius d and d+d ₀ Whether there are other ridge points in the circle range, if not, the ridge point is the center of the circle, and the ridge point of the circle range formed by the radius d is deleted, wherein d and d ₀ are distance constants. The vascular ridge extraction device of claim 8 wherein said enhancement module enhances blood vessels in the three-dimensional angiographic image by a first multi-scale vascular enhancement function as follows:

v ₀ (s) is the first multi-scale vascular enhancement function,

R _A , R _B and S are measure functions, R _{A is} used to distinguish between sheet and line structures, R _{B is} used to distinguish between point structures and line structures, S is used to distinguish background pixels, and α, β and c are threshold values. , λ ₁ , λ ₂ and λ ₃ are three eigenvalues of the Hessian matrix H, and satisfy |λ ₁ | ≤ | λ ₂ | ≤ | λ ₃ |, D is the dimension of the image; The vascular target in the two-dimensional angiography image is enhanced by a second multi-scale vascular enhancement function as follows:

Where v ₀ '(s) is a second multi-scale vascular enhancement function,

R _B ' and S' are the measure functions, R _B ' is used to distinguish the point structure and the line structure, S' is used to distinguish the background pixels, β and c are the threshold values, and λ ₁ and λ ₂ are the two of the Hessian matrix H The feature value, and satisfies |λ ₁ | ≤ | λ ₂ |, D is the dimension of the image. The vascular ridge extraction device according to claim 8, wherein the gradient vector flow field calculation module is specifically configured to: Solve the Euler equations of the gradient vector flow field model as follows:

The iterative formula of the two components of the gradient vector flow field with respect to time after the enhanced two-dimensional angiography image is:

Where n is the number of iterations, f _x (x, y) and f _y (x, y) are the values of the vessel edge function f(x, y) in the x and y directions of the enhanced two-dimensional angiography image, u( x, y), v(x, y) are the two components of the gradient vector flow field V(x, y) = (u(x, y), v(x, y)), and (x, y) is enhanced The coordinates of the pixel points of the posterior two-dimensional angiography image,

For the laplacian operator, μ is the control parameter; Similarly, the iterative formula for the three components of the gradient vector flow field of the enhanced three-dimensional angiography image is:

Where n is the number of iterations, f _x (x, y, z), f _y (x, y, z) and f _z (x, y, z) are the vascular edge functions of the enhanced three-dimensional angiography image f(x , y, z) values in the x, y, and z directions, (x, y, z) are the coordinates of the pixel points of the enhanced three-dimensional angiography image,

For the laplacian operator, μ is the control parameter. The vascular ridge extraction device according to claim 8, wherein the ridge detection module is specifically configured to: For the x-axis direction of the two-dimensional angiography image, it is determined whether a certain pixel point satisfies the following formula, and if satisfied, the pixel point is a ridge point in the x-axis direction:

Where (i, j) is the coordinates of a certain pixel point, (i-1, j) and (i+1, j) are (i, j) two adjacent pixel points in the x-axis direction, V _x ^{( i, j)} , V _x ^{(i-1, j)} , V _x ^{(i+1, j)} are (i, j), (i-1, j), (i+1, j) pixel points, respectively The components of the vector in the x-axis direction, V _y ^(i,j) , V _y ^(i-1,j) , V _y ^(i+1,j) are (i,j), (i-1,j), respectively. a component of the (i+1,j) pixel at the y-axis direction, and T ₀ is a threshold; For the y-axis direction of the two-dimensional angiography image, it is determined whether a certain pixel point satisfies the following formula, and if satisfied, the pixel point is a ridge point in the y-axis direction:

Where (i, j) is the coordinates of a certain pixel point, (i, j-1) and (i, j+1) are (i, j) two adjacent pixel points in the direction of the axis y, V _x ^{( i, j)} , V _x ^{(i, j-1)} , V _x ^{(i, j+1)} are (i, j), (i, j-1), (i, j+1) pixel points, respectively The component of the vector in the x-axis direction, V _y ^(i,j) , V _y ^(i-1,j) , V _y ^(i,j+1) are (i,j), (i,j-1) The component of the (i, j+1) pixel at the y-axis direction, and T ₀ is the threshold. The vascular ridge extraction device according to claim 8, wherein the ridge detection module is specifically configured to: For the x-axis direction of the three-dimensional angiography image, it is determined whether a certain pixel point satisfies the following formula, and if satisfied, the pixel point is a ridge point in the x-axis direction:

Where (i, j, k) is the coordinates of a certain pixel point; (i-1, j, k) and (i+1, j, k) are two (i, j, k) in the x-axis direction Neighboring pixels; V _x ^{(i, j, k)} , V _x ^{(i-1, j, k)} , V _x ^{(i+1, j, k)} are (i, j, k), (i -1, j, k), (i+1, j, k) components of the vector in the x-axis direction; V _y ^{(i, j, k)} , V _y ^{(i-1, j, k)} , V _y ^{(i+1, j, k)} are the components of the vector at the (i, j, k), (i-1, j, k), (i+1, j, k) pixel points in the y-axis direction, respectively. ;V _z ^(i,j,k) , V _z ^(i-1,j,k) , V _z ^(i+1,j,k) are (i,j,k), (i-1,j, respectively , k), (i+1, j, k) a component of the vector at the pixel point in the z-axis direction; T ₀ is a threshold; For the y-axis direction of the three-dimensional angiography image, it is determined whether a certain pixel point satisfies the following formula, and if satisfied, the pixel point is a ridge point in the y-axis direction:

Where (i, j, k) is the coordinates of a certain pixel point; (i, j-1, k) and (i, j+1, k) are two (i, j, k) in the y-axis direction Neighboring pixels; V _x ^{(i, j, k)} , V _x ^{(i, j-1, k)} , V _x ^{(i, j+1, k)} are (i, j, k), (i , j-1, k), (i, j+1, k) the component of the vector in the x-axis direction; V _y ^{(i, j, k)} , V _y ^{(i, j-1, k)} , V _y ^{(i, j+1, k)} are the components of the vector at the (i, j, k), (i, j-1, k), (i, j+1, k) pixel points in the y-axis direction, respectively. ; V _z ^{(i, j, k)} , V _z ^{(i, j-1, k)} , V _z ^{(i, j+1, k)} are (i, j, k), (i, j-1, respectively) , k), (i, j+1, k) a component of the vector at the pixel point in the z-axis direction; T ₀ is a threshold; For the z-axis direction of the three-dimensional angiography image, it is determined whether a certain pixel point satisfies the following formula, and if satisfied, the pixel point is a ridge point in the z-axis direction:

Where (i, j, k) is the coordinates of a certain pixel point; (i, j, k-1) and (i, j, k+1) are two (i, j, k) in the z direction Neighboring pixels; V _x ^{(i, j, k)} , V _x ^{(i, j, k-1)} , V _x ^{(i, j, k+1)} are (i, j, k), (i, respectively j, k-1), (i, j, k+1) the component of the vector at the x-axis direction; V _y ^{(i, j, k)} , V _y ^{(i, j, k-1)} , V _y ^{(i, j, k+1)} are components of the vector at the (i, j, k), (i, j, k-1), (i, j, k+1) pixel points in the y-axis direction, respectively; V _z ^(i,j,k) , V _z ^(i,j,k-1) , V _z ^(i,j,k+1) are (i,j,k), (i,j,k-, respectively) 1), the component of the vector at the (i, j, k+1) pixel point in the z-axis direction; T ₀ is a threshold value.

Images