CN107392891B - Blood vessel tree extraction method, device, equipment and storage medium - Google Patents

Abstract

Translated fromChinese

本发明适用医学图像处理技术领域，提供了一种血管树提取方法、装置、设备及存储介质，该方法包括：将不连续的血管脊线作为初始化的血管曲线；在对所述血管曲线进行迭代的过程中，利用所述迭代后的血管曲线的端点，获取目标曲线的试探点，并根据夹角余弦阈值，从所述试探点中确定所述目标曲线的候选点；通过计算连接矢量与各所述候选点的切向量夹角的最小值，在所述候选点中确定所述目标曲线的融合点；将当前血管曲线的端点与所述目标曲线中融合点的一侧曲线进行融合连接，构成新的血管曲线。本发明提高血管树提取的有效性及精确度，对心血管介入手术中具有重要的临床意义。

The present invention is applicable to the technical field of medical image processing, and provides a blood vessel tree extraction method, device, equipment and storage medium. The method includes: taking discontinuous blood vessel ridge lines as an initialized blood vessel curve; In the process, use the endpoint of the iterative blood vessel curve to obtain the test point of the target curve, and according to the cosine threshold of the included angle, determine the candidate point of the target curve from the test points; The minimum value of the angle between the tangent vectors of the candidate points, and the fusion point of the target curve is determined in the candidate points; the endpoint of the current blood vessel curve is fused and connected to a side curve of the fusion point in the target curve, Constitute a new vascular curve. The invention improves the effectiveness and accuracy of vascular tree extraction, and has important clinical significance in cardiovascular interventional operations.

Description

Translated fromChinese

血管树提取方法、装置、设备及存储介质Vascular tree extraction method, device, equipment and storage medium

技术领域technical field

本发明属于医学图像处理技术领域，尤其涉及一种血管树提取方法、装置、设备及存储介质。The invention belongs to the technical field of medical image processing, and in particular relates to a blood vessel tree extraction method, device, equipment and storage medium.

背景技术Background technique

冠心病严重威胁着人类的健康，随着CT(Computed Tomography，电子计算机断层扫描)成像及其血管造影技术的高度发展，医学图像后处理技术在辅助医生进行诊断治疗中发挥越来越重要的作用。医院影像科采用CTA(CT血管造影术)进行冠心病的诊断时，CT设备捆绑的计算机辅助诊疗软件可以帮助医生判断血管损伤，然而诊断效率低和漏诊等问题依然是不可避免的问题。Coronary heart disease is a serious threat to human health. With the high development of CT (Computed Tomography) imaging and its angiography technology, medical image post-processing technology plays an increasingly important role in assisting doctors in diagnosis and treatment. . When the hospital imaging department uses CTA (CT angiography) for the diagnosis of coronary heart disease, the computer-aided diagnosis and treatment software bundled with the CT equipment can help doctors to determine the vascular damage. However, problems such as low diagnostic efficiency and missed diagnosis are still inevitable problems.

发明内容SUMMARY OF THE INVENTION

本发明的目的在于提供一种血管树提取方法、装置、设备及存储介质，旨在解决由于现有技术无法提供一种有效的血管树提取方法，导致血管树提取不精确、效率低的问题。The purpose of the present invention is to provide a vascular tree extraction method, device, equipment and storage medium, aiming to solve the problems of inaccuracy and low efficiency of vascular tree extraction due to the inability to provide an effective vascular tree extraction method in the prior art.

一方面，本发明提供了一种血管树提取方法，所述方法包括下述步骤：In one aspect, the present invention provides a method for extracting a vascular tree, the method comprising the following steps:

将不连续的血管脊线作为初始化的血管曲线；Take the discontinuous blood vessel ridge line as the initial blood vessel curve;

在对所述血管曲线进行迭代的过程中，利用所述迭代后的血管曲线的端点，获取目标曲线的试探点，并根据夹角余弦阈值，从所述试探点中确定所述目标曲线的候选点；In the process of iterating the blood vessel curve, the end points of the iterative blood vessel curve are used to obtain a tentative point of the target curve, and according to the cosine threshold of the included angle, the candidate of the target curve is determined from the tentative points. point;

通过计算连接矢量与各所述候选点的切向量夹角的最小值，在所述候选点中确定所述目标曲线的融合点，所述连接矢量为所述迭代后的血管曲线的端点与所述候选点的连接矢量；By calculating the minimum value of the angle between the connection vector and the tangent vector of each candidate point, the fusion point of the target curve is determined in the candidate points, and the connection vector is the endpoint of the iterative blood vessel curve and the selected point. the connection vector of the candidate points;

将当前血管曲线的端点与所述目标曲线中融合点的一侧曲线进行融合连接，构成新的血管曲线。The end point of the current blood vessel curve is fused and connected with the curve on one side of the fusion point in the target curve to form a new blood vessel curve.

另一方面，本发明提供了一种血管树提取装置，所述装置包括：In another aspect, the present invention provides a vascular tree extraction device, the device comprising:

初始化单元，用于将不连续的血管脊线作为初始化的血管曲线；The initialization unit is used to use the discontinuous blood vessel ridge line as the initialized blood vessel curve;

候选点确定单元，用于在对所述血管曲线进行迭代的过程中，利用所述迭代后的血管曲线的端点，获取目标曲线的试探点，并根据夹角余弦阈值，从所述试探点中确定所述目标曲线的候选点；a candidate point determination unit, configured to use the endpoints of the iterated blood vessel curve to obtain a tentative point of the target curve in the process of iterating the blood vessel curve, and obtain a tentative point from the tentative point according to the cosine threshold of the included angle determining candidate points of the target curve;

融合点确定单元，用于通过计算连接矢量与各所述候选点的切向量夹角的最小值，在所述候选点中确定所述目标曲线的融合点，所述连接矢量为所述迭代后的血管曲线的端点与所述候选点的连接矢量；以及The fusion point determination unit is used to determine the fusion point of the target curve in the candidate points by calculating the minimum value of the angle between the connection vector and the tangent vector of each of the candidate points, and the connection vector is after the iteration The connection vector of the endpoint of the blood vessel curve and the candidate point; and

融合连接单元，用于将当前血管曲线的端点与所述目标曲线中融合点的一侧曲线进行融合连接，构成新的血管曲线。The fusion connection unit is used for fusion connection between the end point of the current blood vessel curve and the one side curve of the fusion point in the target curve to form a new blood vessel curve.

另一方面，本发明还提供了一种血管树提取设备，包括存储器、处理器以及存储在所述存储器中并可在所述处理器上运行的计算机程序，所述处理器执行所述计算机程序时实现如血管树提取方法的步骤。In another aspect, the present invention also provides a vessel tree extraction device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the computer program When implementing steps such as the vascular tree extraction method.

另一方面，本发明还提供了一种计算机可读存储介质，所述计算机可读存储介质存储有计算机程序，所述计算机程序被处理器执行时实现如血管树提取方法的步骤。In another aspect, the present invention also provides a computer-readable storage medium, where the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, implements steps such as a method for extracting a blood vessel tree.

本发明将初始化的血管曲线与目标曲线进行融合，在目标曲线获取试探点，并根据夹角余弦阈值确定目标曲线的候选点，利用切向量夹角的最小值确定最终与目标曲线进行融合的融合点，构成新的血管曲线，提高了血管树提取的有效性及精确度，精确高效地提取血管分支和血管树是血管可视化、计算机辅助诊断、介入路径规划、手术导航的关键，对心血管介入手术中具有重要的临床意义。The present invention fuses the initialized blood vessel curve with the target curve, obtains the test point on the target curve, determines the candidate point of the target curve according to the cosine threshold of the included angle, and uses the minimum value of the included angle of the tangent vector to determine the final fusion with the target curve. It forms a new vascular curve, which improves the effectiveness and accuracy of vascular tree extraction. Accurate and efficient extraction of vascular branches and vascular tree is the key to vascular visualization, computer-aided diagnosis, interventional path planning, and surgical navigation. Surgery has important clinical significance.

附图说明Description of drawings

图1是本发明实施例一提供的血管树提取方法的实现流程图；Fig. 1 is the realization flow chart of the blood vessel tree extraction method provided by the first embodiment of the present invention;

图2是本发明实施例提供的血管曲线融合过程的示意图；2 is a schematic diagram of a blood vessel curve fusion process provided by an embodiment of the present invention;

图3是本发明实施例二提供的血管树提取方法的实现流程图；Fig. 3 is the realization flow chart of the blood vessel tree extraction method provided by the second embodiment of the present invention;

图4是本发明实施例三提供的血管树提取装置的结构示意图；4 is a schematic structural diagram of a blood vessel tree extraction device provided in Embodiment 3 of the present invention;

图5是本发明实施例四提供的血管树提取设备的结构示意图。FIG. 5 is a schematic structural diagram of a blood vessel tree extraction device provided in Embodiment 4 of the present invention.

具体实施方式Detailed ways

为了使本发明的目的、技术方案及优点更加清楚明白，以下结合附图及实施例，对本发明进行进一步详细说明。应当理解，此处所描述的具体实施例仅仅用以解释本发明，并不用于限定本发明。In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, but not to limit the present invention.

以下结合具体实施例对本发明的具体实现进行详细描述：The specific implementation of the present invention is described in detail below in conjunction with specific embodiments:

实施例一：Example 1:

图1示出了本发明实施例一提供的血管树提取方法的实现流程，为了便于说明，仅示出了与本发明实施例相关的部分，详述如下：FIG. 1 shows the implementation process of the method for extracting a blood vessel tree provided by the first embodiment of the present invention. For the convenience of description, only the part related to the embodiment of the present invention is shown, and the details are as follows:

在步骤S101中，将不连续的血管脊线作为初始化的血管曲线。In step S101, the discontinuous blood vessel ridge line is used as the initialized blood vessel curve.

在本发明实施例中，获取CTA造影中的主动脉、冠状动脉的根点、以及心血管增强后的图像，将不连续的血管脊线作为初始化的血管曲线。In the embodiment of the present invention, the aorta, the root point of the coronary artery, and the enhanced cardiovascular image in the CTA angiography are acquired, and the discontinuous blood vessel ridge line is used as the initialized blood vessel curve.

进一步地，获取血管曲线内各点的位置坐标及切向量，并对血管曲线标记端点属性及曲线属性。Further, the position coordinates and tangent vectors of each point in the blood vessel curve are acquired, and the endpoint attributes and curve attributes are marked on the blood vessel curve.

具体地，将不连续的血管脊线作为初始化的血管曲线S_N，S_N＝{X,V,P,T}_N，其中，N表示血管曲线的总数，X_n，V_n分别表示第n条血管曲线内的各点的位置坐标及切向量，P_n表示第n条血管曲线内的端点属性，及T_n表示第n条血管曲线的曲线属性，在初始化时，需要对血管曲线标记端点属性及曲线属性。Specifically, the discontinuous blood vessel ridge line is used as the initialized blood vessel curve S_N , S_N ={X,V,P,T}_N , where N represents the total number of blood vessel curves, and X_n and V_n represent the nth blood vessel curve respectively. The position coordinates and tangent vector of each point in the blood vessel curve, P_n represents the endpoint property of the nth blood vessel curve, and T_n represents the curve property of the nth blood vessel curve. During initialization, the endpoints of the blood vessel curve need to be marked properties and curve properties.

具体地，初始化时，连接主动脉根点的血管曲线分别属于左、右冠状动脉，英文缩写即：LCA和RCA，被分别标记为曲线属性为T＝1及T＝2，其它的血管曲线的曲线属性都为T＝0，左冠状动脉的端点属性为P＝[0；-1]，右冠状动脉的端点属性为P＝[0；-1]，其他的血管曲线的端点属性为P＝[-1；-1]。Specifically, during initialization, the blood vessel curves connecting the aortic root points belong to the left and right coronary arteries, respectively, the English abbreviations are: LCA and RCA, which are marked as curve attributes T=1 and T=2 respectively, and the other blood vessel curves are The curve attributes are all T=0, the endpoint property of the left coronary artery is P=[0;-1], the endpoint property of the right coronary artery is P=[0;-1], and the endpoint properties of other blood vessel curves are P= [-1;-1].

在步骤S102中，在对血管曲线进行迭代的过程中，利用迭代后的血管曲线的端点，获取目标曲线的试探点，并根据夹角余弦阈值，从试探点中确定目标曲线的候选点。In step S102 , in the process of iterating the blood vessel curve, the end points of the iterated blood vessel curve are used to obtain the test points of the target curve, and according to the cosine threshold of the included angle, the candidate points of the target curve are determined from the test points.

在本发明实施例中，在对血管曲线进行迭代的过程中，获取迭代后的血管曲线的端点，将与当前的血管曲线待融合的血管曲线作为目标曲线，当然，目标曲线可以不唯一，获取目标曲线的试探点，为了缩小查找范围，可以将该试探点根据血管曲线的端点的预设范围内进行获取，根据夹角余弦阈值，从试探点中确定目标曲线的候选点。In the embodiment of the present invention, in the process of iterating the blood vessel curve, the endpoints of the iterative blood vessel curve are obtained, and the blood vessel curve to be fused with the current blood vessel curve is used as the target curve. Of course, the target curve may not be unique. The tentative point of the target curve, in order to narrow the search range, the tentative point can be obtained according to the preset range of the endpoint of the blood vessel curve, and the candidate point of the target curve is determined from the tentative points according to the cosine threshold of the included angle.

进一步地，在迭代后的血管曲线的端点的预设范围内获取试探点，计算端点的方向矢量与端点到试探点的连接方向矢量之间的方向夹角的余弦值；Further, a tentative point is obtained within the preset range of the endpoint of the iterative blood vessel curve, and the cosine value of the included direction angle between the direction vector of the endpoint and the direction vector connecting the endpoint to the tentative point is calculated;

当余弦值满足夹角余弦阈值时，确定当前试探点为目标曲线的候选点。When the cosine value satisfies the cosine threshold value of the included angle, it is determined that the current test point is a candidate point of the target curve.

具体地，设第n条血管曲线的第t步迭代的进化结果的为

其左右端点为分别为

和

与当前血管曲线

发生融合的目标曲线表示为S_m，该目标曲线S_m可能不止一条，血管曲线融合过程如下：首先，在

的左右端点

和

的预设范围δ内获取试探点，其中，预设范围δ可以为δ＝2r_max，r_max为血管最大半径，在该预设范围δ内获取可能存在的试探点

当目标曲线S_m不止一条时，每个试探点的对应于所在脊线序号，如m₁,m₂,m₃等进行区分，S_m中的候选点为

其与

端点的连接矢量表示为

则候选点的确定需满足如下夹角余弦阈值条件：Specifically, let the evolution result of the t-th iteration of the n-th blood vessel curve be

Its left and right endpoints are respectively

and

with the current vessel curve

The fusion target curve is denoted as S_m , there may be more than one target curve S_m , and the fusion process of the blood vessel curves is as follows: First, in

the left and right endpoints of

and

The tentative points are acquired within the preset range δ of , where the preset range δ can be δ=2r_max , r_max is the maximum radius of the blood vessel, and possible tentative points are acquired within the preset range δ

When there is more than one target curve S_m , each test point corresponds to the ridge line number, such as m₁ , m₂ , m₃ , etc. to distinguish, and the candidate points in S_m are

its and

The connection vector of the endpoints is represented as

The determination of candidate points must meet the following cosine threshold conditions:

其中，

为当前端点

的方向矢量；

为当前端点到候选点

的连接矢量，

为该连接矢量的长度；上述两个矢量的方向夹角余弦值满足

的点被作为候选点，一般选择T_θ＝0，如图2，血管曲线端点

为当前血管曲线

的端点，目标曲线S_m上的点

为试探点，而

为满足夹角余弦阈值条件的候选点。in,

is the current endpoint

direction vector;

for the current endpoint to the candidate point

the connection vector of ,

is the length of the connection vector; the cosine of the angle between the directions of the two vectors above satisfies

The points of , are used as candidate points, generally choose T_θ = 0, as shown in Figure 2, the endpoint of the blood vessel curve

is the current vessel curve

The endpoint of , the point on the target curve S_m

as a tentative point, and

The candidate points that satisfy the cosine threshold condition of the included angle.

在步骤S103中，通过计算连接矢量与各候选点的切向量夹角的最小值，在候选点中确定目标曲线的融合点，该连接矢量为迭代后的血管曲线的端点与候选点的连接矢量。In step S103, by calculating the minimum value of the angle between the connection vector and the tangent vector of each candidate point, the fusion point of the target curve is determined in the candidate points, and the connection vector is the connection vector between the endpoint of the iterative blood vessel curve and the candidate point. .

在本发明实施例中，从候选点中确定融合点，利用连接矢量

与目标曲线S_m上各候选点的切向量夹角，夹角最小的连接点作为融合点，即最佳连接点。In the embodiment of the present invention, the fusion point is determined from the candidate points, and the connection vector is used to determine the fusion point.

The angle with the tangent vector of each candidate point on the target curve S_m , the connection point with the smallest angle is used as the fusion point, that is, the best connection point.

具体地，根据夹角余弦阈值条件下产生的多个候选点

中，如图2，

为候选点，为了确定最佳连接点，需利用连接矢量

与目标曲线S_m上各候选点的切向量夹角

夹角最小的连接点作为融合点：Specifically, according to the multiple candidate points generated under the condition of the cosine threshold of the included angle

, as shown in Figure 2,

is a candidate point, in order to determine the best connection point, it is necessary to use the connection vector

The angle with the tangent vector of each candidate point on the target curve S_m

The connection point with the smallest included angle is used as the fusion point:

即为当前迭代后的血管曲线

和目标曲线S_m的融合点，如图2所示

is the blood vessel curve after the current iteration

and the fusion point of the target curve S_m , as shown in Figure 2

在步骤S104中，将当前血管曲线的端点与目标曲线中融合点的一侧曲线进行融合连接，构成新的血管曲线。In step S104, the end point of the current blood vessel curve is fused and connected with the curve on one side of the fusion point in the target curve to form a new blood vessel curve.

在本发明实施例中，融合就是将当前血管曲线

和目标曲线S_m中融合点

的一侧相连接，构成一条新的血管曲线

如图2所示。因为当前血管曲线

和目标曲线S_m是由位置坐标X、切向量V、端点属性P和曲线属性T构成，因此，曲线融合之后依然需要维持该有序的数据结构。In this embodiment of the present invention, fusion is to combine the current blood vessel curve

and the fusion point in the target curve S_m

connected on one side to form a new vascular curve

as shown in picture 2. Because the current vessel curve

The sum target curve S_m is composed of the position coordinate X, the tangent vector V, the endpoint attribute P and the curve attribute T. Therefore, the ordered data structure still needs to be maintained after the curves are fused.

进一步地，融合方式包括：将血管曲线的左端点与目标曲线中融合点的左侧曲线进行融合连接；或Further, the fusion method includes: fusion and connection between the left end point of the blood vessel curve and the left curve of the fusion point in the target curve; or

将血管曲线的左端点与目标曲线中融合点的右侧曲线进行融合连接；或Blend the left endpoint of the vessel curve with the right curve of the blend point in the target curve; or

将血管曲线的右端点与目标曲线中融合点的左侧曲线进行融合连接；或Fusion join the right endpoint of the vessel curve with the left curve of the fusion point in the target curve; or

将血管曲线的右端点与目标曲线中融合点的右侧曲线进行融合连接。Blend the right endpoint of the vessel curve with the right curve of the fusion point in the target curve.

具体地，以将血管曲线的左端点与目标曲线中融合点的右侧曲线进行融合连接的融合方式为例，新的血管曲线

中各分量描述为：Specifically, taking the fusion method of merging and connecting the left endpoint of the blood vessel curve and the right curve of the fusion point in the target curve as an example, the new blood vessel curve

The components are described as:

其中，

表示融合后的坐标序列

由当前血管曲线

的坐标序列

和目标曲线

的反向坐标序列{x_j,j＝j′,...,1}_m构成；

表示融合后的方向矢量序列

由融合后的坐标序列

的差分直接计算得到；

表示融合后曲线的端点坐标

分别由曲线

和

的端点状态

和

得到；

表示新的曲线属性等于融合前两个曲线属性中的最大值，可以这样理解：除了连接主动脉根点的两条初始血管曲线(分别属于左、右冠状动脉)被分别标记为属性值T＝1和T＝2，其它所有血管曲线属性都为T＝0，逐条曲线的进化过程中：标记为T＝0的血管曲线和T＝1或T＝2的血管曲线融合，将分别产生T＝1或T＝2的同类血管，而T＝1和T＝2两种曲线属性的血管之间不能发生融合连接，从而各自产生左、右冠状动脉树。in,

Represents the fused coordinate sequence

by the current vessel curve

sequence of coordinates

and target curve

The reverse coordinate sequence {x_j ,j=j′,...,1}_m constitutes;

Represents the fused sequence of direction vectors

By the fused coordinate sequence

The difference of can be directly calculated;

Indicates the coordinates of the endpoints of the fused curve

respectively by the curve

and

the endpoint status of

and

get;

Indicates that the new curve attribute is equal to the maximum value of the two curve attributes before fusion, which can be understood as follows: In addition to the two initial blood vessel curves connecting the aortic root points (belonging to the left and right coronary arteries, respectively), they are marked as attribute values T= 1 and T=2, all other blood vessel curve attributes are T=0. During the evolution process of each curve: the blood vessel curve marked with T=0 and the blood vessel curve marked with T=1 or T=2 are merged, and T= 1 or T=2 of the same kind of blood vessels, while T=1 and T=2 two kinds of blood vessels with curve properties cannot be fused and connected, so that the left and right coronary artery trees are generated respectively.

在本发明实施例中，将初始化的血管曲线与目标曲线进行融合，在目标曲线获取试探点，并根据夹角余弦阈值确定目标曲线的候选点，利用切向量夹角的最小值确定最终与目标曲线进行融合的融合点，构成新的血管曲线，提高了血管树提取的有效性及精确度，精确高效地提取血管分支和血管树是血管可视化、计算机辅助诊断、介入路径规划、手术导航的关键，对心血管介入手术中具有重要的临床意义。In the embodiment of the present invention, the initialized blood vessel curve and the target curve are fused, the tentative points are obtained on the target curve, and the candidate points of the target curve are determined according to the cosine threshold of the included angle, and the minimum value of the included angle of the tangent vector is used to determine the final value of the target curve. The fusion point of the curve fusion forms a new blood vessel curve, which improves the effectiveness and accuracy of the blood vessel tree extraction. Accurate and efficient extraction of blood vessel branches and blood vessel trees is the key to blood vessel visualization, computer-aided diagnosis, interventional path planning, and surgical navigation. , has important clinical significance in cardiovascular interventional surgery.

实施例二：Embodiment 2:

图3示出了本发明实施例二提供的血管树提取方法的实现流程，为了便于说明，仅示出了与本发明实施例相关的部分，详述如下：FIG. 3 shows the implementation process of the method for extracting a blood vessel tree provided by the second embodiment of the present invention. For the convenience of description, only the part related to the embodiment of the present invention is shown, and the details are as follows:

在步骤S201中，将不连续的血管脊线作为初始化的血管曲线。In step S201, the discontinuous blood vessel ridge line is used as the initialized blood vessel curve.

在步骤S202中，在对血管曲线进行迭代的过程中，利用迭代后的血管曲线的端点，获取目标曲线的试探点，并根据夹角余弦阈值，从试探点中确定目标曲线的候选点。In step S202 , in the process of iterating the blood vessel curve, the end points of the iterated blood vessel curve are used to obtain the tentative points of the target curve, and the candidate points of the target curve are determined from the tentative points according to the cosine threshold of the included angle.

在步骤S203中，通过计算连接矢量与各候选点的切向量夹角的最小值，在候选点中确定目标曲线的融合点，该连接矢量为迭代后的血管曲线的端点与候选点的连接矢量。In step S203, by calculating the minimum value of the angle between the connection vector and the tangent vector of each candidate point, the fusion point of the target curve is determined in the candidate points, and the connection vector is the connection vector between the endpoint of the iterative blood vessel curve and the candidate point .

在步骤S204中，将当前血管曲线的端点与目标曲线中融合点的一侧曲线进行融合连接，构成新的血管曲线。In step S204, the end point of the current blood vessel curve is fused and connected with the curve on one side of the fusion point in the target curve to form a new blood vessel curve.

在步骤S205中，获取新的血管曲线上逐点的局部测度值，将逐点的局部测度值构成测度值曲线，查找测度值曲线的波谷，截去波谷所在点的右侧曲线。In step S205, the point-by-point local measurement value on the new blood vessel curve is obtained, the point-by-point local measurement value is formed into a measurement value curve, the trough of the measurement value curve is searched, and the right curve of the point where the trough is located is truncated.

在步骤S206中，在新的血管曲线上，计算每一点相对于距离当前点为预设曲线长度的所在点的切线方向的变化程度，保留变化程度大于预设变化程度的血管曲线段。In step S206, on the new blood vessel curve, calculate the change degree of each point relative to the tangential direction of the point where the distance from the current point is the preset curve length, and retain the blood vessel curve segment with the change degree greater than the preset change degree.

在本发明实施例中，因为血管曲线融合策略是在初始N条血管曲线的基础上逐条完成的，血管曲线条数不变的情况下将会造成很多重叠的曲线，这是同一条血管路径的多条血管中心线的冗余效果，同样，最终的血管曲线集合S_N＝{X,V,P,T}_N包含了多个血管树(心脏只有2个冠状动脉树：一为LCA树，二是RCA树)；另外，在处理过程中也提取出少许其它血管，例如心脏静脉，及组织边缘，例如心脏房室，因此，需要进一步对血管树进行修剪。In the embodiment of the present invention, because the blood vessel curve fusion strategy is completed one by one on the basis of the initial N blood vessel curves, many overlapping curves will be caused when the number of blood vessel curves remains unchanged, which is the result of the same blood vessel path. Redundancy effect of multiple vascular centerlines. Similarly, the final vascular curve set S_N = {X, V, P, T}_N contains multiple vascular trees (the heart has only 2 coronary trees: one is the LCA tree, The second is the RCA tree); in addition, a few other blood vessels, such as cardiac veins, and tissue edges, such as cardiac atrioventricular, are also extracted during processing, so the vessel tree needs to be further pruned.

对初始化两条沿左右冠脉旅行的路径：分别从两条血管树根部的两个血管树的冠脉开口位置(起始点分别对应于x₀₁和x₀₂的位置)出发，向两个血管树末梢旅行，将遍历每一条血管中心线分支，对途经的各个血管分支按照以下准则进行分析：To initialize two paths that travel along the left and right coronary arteries: starting from the coronary opening positions of the two vascular trees at the root of the two vascular trees (the starting points correspond to the positions of x₀₁ and x₀₂ , respectively), to the two vascular trees. Peripheral travel will traverse each blood vessel centerline branch, and analyze each blood vessel branch along the way according to the following criteria:

获取新的血管曲线上逐点的局部测度值，将逐点的局部测度值构成测度值曲线，查找测度值曲线的波谷，截去波谷所在点的右侧曲线。Obtain the point-by-point local measurement value on the new blood vessel curve, form the point-by-point local measurement value into a measurement value curve, find the trough of the measurement value curve, and truncate the right curve of the point where the trough is located.

具体地，在血管增强图像

中测量每一条血管分支{X}_n上逐点的局部测度值，局部测度值计算公式为：Specifically, in the blood vessel enhanced image

Measure the point-by-point local measurement value of each blood vessel branch {X}_n in , and the calculation formula of the local measurement value is:

以上公式代表第n条血管曲线的第i点(x_i,i＝1,...,M_n)的局部测度值E_n(x_i)，采用相邻脊点区间

内强度的均值和方差计算，

为曲线长度。The above formula represents the local measurement value En (x_i ) of the i-_th point (x_i ,i=1,...,M_n ) of the n-th blood vessel curve, using the interval between adjacent ridge points

mean and variance calculation of inner strength,

is the length of the curve.

在新的血管曲线上，计算每一点相对于距离当前点为预设曲线长度的所在点的切线方向的变化程度，保留变化程度大于预设变化程度的血管曲线段。On the new blood vessel curve, the degree of change of each point in the tangential direction relative to the point where the distance from the current point is the preset curve length is calculated, and the blood vessel curve segments with the degree of change greater than the preset degree of change are retained.

具体地，在新的血管曲线上，在预设曲线长度L内，测量第i点相对于第i+L点的切线方向的变化程度，用预设变化程度

保留变化程度大于预设变化程度的血管曲线段，通常取

L＝30。Specifically, on the new blood vessel curve, within the preset curve length L, measure the degree of change in the tangential direction of the i-th point relative to the i+L-th point, and use the preset change degree

Retain the blood vessel curve segment with the change degree greater than the preset change degree, usually take the

L=30.

在本发明实施例中，将初始化的血管曲线与目标曲线进行融合，在目标曲线获取试探点，并根据夹角余弦阈值确定目标曲线的候选点，利用切向量夹角的最小值确定最终与目标曲线进行融合的融合点，构成新的血管曲线，提高了血管树提取的有效性及精确度，精确高效地提取血管分支和血管树是血管可视化、计算机辅助诊断、介入路径规划、手术导航的关键，对心血管介入手术中具有重要的临床意义。In the embodiment of the present invention, the initialized blood vessel curve and the target curve are fused, the tentative points are obtained on the target curve, and the candidate points of the target curve are determined according to the cosine threshold of the included angle, and the minimum value of the included angle of the tangent vector is used to determine the final value of the target curve. The fusion point of the curve fusion forms a new blood vessel curve, which improves the effectiveness and accuracy of the blood vessel tree extraction. Accurate and efficient extraction of blood vessel branches and blood vessel trees is the key to blood vessel visualization, computer-aided diagnosis, interventional path planning, and surgical navigation. , has important clinical significance in cardiovascular interventional surgery.

实施例三：Embodiment three:

图4示出了本发明实施例三提供的血管树提取装置的结构示意图，为了便于说明，仅示出了与本发明实施例相关的部分，血管树提取装置包括：FIG. 4 shows a schematic structural diagram of a blood vessel tree extraction device provided by Embodiment 3 of the present invention. For convenience of description, only the parts related to the embodiment of the present invention are shown. The blood vessel tree extraction device includes:

初始化单元31，用于将不连续的血管脊线作为初始化的血管曲线。Theinitialization unit 31 is configured to use the discontinuous blood vessel ridge line as the initialized blood vessel curve.

在本发明实施例中，获取CTA造影中的主动脉、冠状动脉的根点、以及心血管增强后的图像，将不连续的血管脊线作为初始化的血管曲线。In the embodiment of the present invention, the aorta, the root point of the coronary artery, and the enhanced cardiovascular image in the CTA angiography are acquired, and the discontinuous blood vessel ridge line is used as the initialized blood vessel curve.

进一步地，获取血管曲线内各点的位置坐标及切向量，并对血管曲线标记端点属性及曲线属性。Further, the position coordinates and tangent vectors of each point in the blood vessel curve are acquired, and the endpoint attributes and curve attributes are marked on the blood vessel curve.

具体地，将不连续的血管脊线作为初始化的血管曲线S_N，S_N＝{X,V,P,T}_N，其中，N表示血管曲线的总数，X_n，V_n分别表示第n条血管曲线内的各点的位置坐标及切向量，P_n表示第n条血管曲线内的端点属性，及T_n表示第n条血管曲线的曲线属性，在初始化时，需要对血管曲线标记端点属性及曲线属性。Specifically, the discontinuous blood vessel ridge line is used as the initialized blood vessel curve S_N , S_N ={X,V,P,T}_N , where N represents the total number of blood vessel curves, and X_n and V_n represent the nth blood vessel curve respectively. The position coordinates and tangent vector of each point in the blood vessel curve, P_n represents the endpoint property of the nth blood vessel curve, and T_n represents the curve property of the nth blood vessel curve. During initialization, the endpoints of the blood vessel curve need to be marked properties and curve properties.

具体地，初始化时，连接主动脉根点的血管曲线分别属于左、右冠状动脉，英文缩写即：LCA和RCA，被分别标记为曲线属性为T＝1及T＝2，其它的血管曲线的曲线属性都为T＝0，左冠状动脉的端点属性为P＝[0；-1]，右冠状动脉的端点属性为P＝[0；-1]，其他的血管曲线的端点属性为P＝[-1；-1]。Specifically, during initialization, the blood vessel curves connecting the aortic root points belong to the left and right coronary arteries, respectively, the English abbreviations are: LCA and RCA, which are marked as curve attributes T=1 and T=2 respectively, and the other blood vessel curves are The curve attributes are all T=0, the endpoint property of the left coronary artery is P=[0;-1], the endpoint property of the right coronary artery is P=[0;-1], and the endpoint properties of other blood vessel curves are P= [-1;-1].

候选点确定单元32，用于在对血管曲线进行迭代的过程中，利用迭代后的血管曲线的端点，获取目标曲线的试探点，并根据夹角余弦阈值，从试探点中确定目标曲线的候选点。The candidatepoint determination unit 32 is configured to, in the process of iterating the blood vessel curve, use the endpoints of the iterated blood vessel curve to obtain the test point of the target curve, and determine the candidate of the target curve from the test points according to the cosine threshold of the included angle point.

在本发明实施例中，在对血管曲线进行迭代的过程中，获取迭代后的血管曲线的端点，将与当前的血管曲线待融合的血管曲线作为目标曲线，当然，目标曲线可以不唯一，获取目标曲线的试探点，为了缩小查找范围，可以将该试探点根据血管曲线的端点的预设范围内进行获取，根据夹角余弦阈值，从试探点中确定目标曲线的候选点。In the embodiment of the present invention, in the process of iterating the blood vessel curve, the endpoints of the iterative blood vessel curve are obtained, and the blood vessel curve to be fused with the current blood vessel curve is used as the target curve. Of course, the target curve may not be unique. The tentative point of the target curve, in order to narrow the search range, the tentative point can be obtained according to the preset range of the endpoint of the blood vessel curve, and the candidate point of the target curve is determined from the tentative points according to the cosine threshold of the included angle.

进一步地，在迭代后的血管曲线的端点的预设范围内获取试探点，计算端点的方向矢量与端点到试探点的连接方向矢量之间的方向夹角的余弦值；Further, a tentative point is obtained within the preset range of the endpoint of the iterative blood vessel curve, and the cosine value of the included direction angle between the direction vector of the endpoint and the direction vector connecting the endpoint to the tentative point is calculated;

当余弦值满足夹角余弦阈值时，确定当前试探点为目标曲线的候选点。When the cosine value satisfies the cosine threshold value of the included angle, it is determined that the current test point is a candidate point of the target curve.

具体地，设第n条血管曲线的第t步迭代的进化结果的为

其左右端点为分别为

和

与当前血管曲线

发生融合的目标曲线表示为S_m，该目标曲线S_m可能不止一条，血管曲线融合过程如下：首先，在

的左右端点

和

的预设范围δ内获取试探点，其中，预设范围δ可以为δ＝2r_max，r_max为血管最大半径，在该预设范围δ内获取可能存在的试探点

当目标曲线S_m不止一条时，每个试探点的对应于所在脊线序号，如m₁,m₂,m₃等进行区分，S_m中的候选点为

其与

端点的连接矢量表示为

则候选点的确定需满足如下夹角余弦阈值条件：Specifically, let the evolution result of the t-th iteration of the n-th blood vessel curve be

Its left and right endpoints are respectively

and

with the current vessel curve

The fusion target curve is denoted as S_m , there may be more than one target curve S_m , and the fusion process of the blood vessel curves is as follows: First, in

the left and right endpoints of

and

The tentative points are acquired within the preset range δ of , where the preset range δ can be δ=2r_max , r_max is the maximum radius of the blood vessel, and possible tentative points are acquired within the preset range δ

When there is more than one target curve S_m , each test point corresponds to the ridge line number, such as m₁ , m₂ , m₃ , etc. to distinguish, and the candidate points in S_m are

its and

The connection vector of the endpoints is represented as

The determination of candidate points must meet the following cosine threshold conditions:

其中，

为当前端点

的方向矢量；

为当前端点到候选点

的连接矢量，

为该连接矢量的长度；上述两个矢量的方向夹角余弦值满足

的点被作为候选点，一般选择T_θ＝0，如图2，血管曲线端点

为当前血管曲线

的端点，目标曲线S_m上的点

为试探点，而

为满足夹角余弦阈值条件的候选点。in,

is the current endpoint

direction vector;

for the current endpoint to the candidate point

the connection vector of ,

is the length of the connection vector; the cosine of the angle between the directions of the two vectors above satisfies

The points of , are used as candidate points, generally choose T_θ = 0, as shown in Figure 2, the endpoint of the blood vessel curve

is the current vessel curve

The endpoint of , the point on the target curve S_m

as a tentative point, and

The candidate points that satisfy the cosine threshold condition of the included angle.

融合点确定单元33，用于通过计算连接矢量与各候选点的切向量夹角的最小值，在候选点中确定目标曲线的融合点，连接矢量为迭代后的血管曲线的端点与所述候选点的连接矢量。The fusionpoint determination unit 33 is used to determine the fusion point of the target curve in the candidate points by calculating the minimum value of the tangent vector angle between the connection vector and each candidate point, and the connection vector is the endpoint of the iterative blood vessel curve and the candidate point. Connection vector of dots.

在本发明实施例中，从候选点中确定融合点，利用连接矢量

与目标曲线S_m上各候选点的切向量夹角，夹角最小的连接点作为融合点，即最佳连接点。In the embodiment of the present invention, the fusion point is determined from the candidate points, and the connection vector is used to determine the fusion point.

The angle with the tangent vector of each candidate point on the target curve S_m , the connection point with the smallest angle is used as the fusion point, that is, the best connection point.

具体地，根据夹角余弦阈值条件下产生的多个候选点

中，如图2，

为候选点，为了确定最佳连接点，需利用连接矢量

与目标曲线S_m上各候选点的切向量夹角

夹角最小的连接点作为融合点：Specifically, according to the multiple candidate points generated under the condition of the cosine threshold of the included angle

, as shown in Figure 2,

is a candidate point, in order to determine the best connection point, it is necessary to use the connection vector

The angle with the tangent vector of each candidate point on the target curve S_m

The connection point with the smallest included angle is used as the fusion point:

即为当前迭代后的血管曲线

和目标曲线S_m的融合点，如图2所示

is the blood vessel curve after the current iteration

and the fusion point of the target curve S_m , as shown in Figure 2

融合连接单元34，用于将当前血管曲线的端点与目标曲线中融合点的一侧曲线进行融合连接，构成新的血管曲线。Thefusion connection unit 34 is used for fusion connection between the end point of the current blood vessel curve and the curve on one side of the fusion point in the target curve to form a new blood vessel curve.

在本发明实施例中，融合就是将当前血管曲线

和目标曲线S_m中融合点

的一侧相连接，构成一条新的血管曲线

如图2所示。因为当前血管曲线

和目标曲线S_m是由位置坐标X、切向量V、端点属性P和曲线属性T构成，因此，曲线融合之后依然需要维持该有序的数据结构。In this embodiment of the present invention, fusion is to combine the current blood vessel curve

and the fusion point in the target curve S_m

connected on one side to form a new vascular curve

as shown in picture 2. Because the current vessel curve

The sum target curve S_m is composed of the position coordinate X, the tangent vector V, the endpoint attribute P and the curve attribute T. Therefore, the ordered data structure still needs to be maintained after the curves are fused.

进一步地，融合方式包括：将血管曲线的左端点与目标曲线中融合点的左侧曲线进行融合连接；或Further, the fusion method includes: fusion and connection between the left end point of the blood vessel curve and the left curve of the fusion point in the target curve; or

将血管曲线的左端点与目标曲线中融合点的右侧曲线进行融合连接；或Blend the left endpoint of the vessel curve with the right curve of the blend point in the target curve; or

将血管曲线的右端点与目标曲线中融合点的左侧曲线进行融合连接；或Fusion join the right endpoint of the vessel curve with the left curve of the fusion point in the target curve; or

将血管曲线的右端点与目标曲线中融合点的右侧曲线进行融合连接。Blend the right endpoint of the vessel curve with the right curve of the fusion point in the target curve.

具体地，以将血管曲线的左端点与目标曲线中融合点的右侧曲线进行融合连接的融合方式为例，新的血管曲线

中各分量描述为：Specifically, taking the fusion method of merging and connecting the left endpoint of the blood vessel curve and the right curve of the fusion point in the target curve as an example, the new blood vessel curve

The components are described as:

其中，

表示融合后的坐标序列

由当前血管曲线

的坐标序列

和目标曲线

的反向坐标序列{x_j,j＝j′,...,1}_m构成；

表示融合后的方向矢量序列

由融合后的坐标序列

的差分直接计算得到；

表示融合后曲线的端点坐标

分别由曲线

和

的端点状态

和

得到；

表示新的曲线属性等于融合前两个曲线属性中的最大值，可以这样理解：除了连接主动脉根点的两条初始血管曲线(分别属于左、右冠状动脉)被分别标记为属性值T＝1和T＝2，其它所有血管曲线属性都为T＝0，逐条曲线的进化过程中：标记为T＝0的血管曲线和T＝1或T＝2的血管曲线融合，将分别产生T＝1或T＝2的同类血管，而T＝1和T＝2两种曲线属性的血管之间不能发生融合连接，从而各自产生左、右冠状动脉树。in,

Represents the fused coordinate sequence

by the current vessel curve

sequence of coordinates

and target curve

The reverse coordinate sequence {x_j ,j=j′,...,1}_m constitutes;

Represents the fused sequence of direction vectors

By the fused coordinate sequence

The difference of can be directly calculated;

Indicates the coordinates of the endpoints of the fused curve

respectively by the curve

and

the endpoint status of

and

get;

Indicates that the new curve attribute is equal to the maximum value of the two curve attributes before fusion, which can be understood as follows: In addition to the two initial blood vessel curves connecting the aortic root points (belonging to the left and right coronary arteries, respectively), they are marked as attribute values T= 1 and T=2, all other blood vessel curve attributes are T=0. During the evolution process of each curve: the blood vessel curve marked with T=0 and the blood vessel curve marked with T=1 or T=2 are merged, and T= 1 or T=2 of the same type of blood vessels, while T=1 and T=2 two kinds of blood vessels of the same kind of curve properties cannot be merged and connected, so that the left and right coronary artery trees are generated respectively.

进一步地，血管树提取装置还包括：曲线截去单元，用于获取新的血管曲线上逐点的局部测度值，将逐点的局部测度值构成测度值曲线，查找测度值曲线的波谷，截去所述波谷所在点的右侧曲线。Further, the blood vessel tree extraction device also includes: a curve truncation unit, used for acquiring point-by-point local measurement values on the new blood vessel curve, forming a measurement value curve from the point-by-point local measurement values, searching for the trough of the measurement value curve, and cutting the curve. Go to the curve to the right of the point where the trough is located.

在本发明实施例中，因为血管曲线融合策略是在初始N条血管曲线的基础上逐条完成的，血管曲线条数不变的情况下将会造成很多重叠的曲线，这是同一条血管路径的多条血管中心线的冗余效果，同样，最终的血管曲线集合S_N＝{X,V,P,T}_N包含了多个血管树(心脏只有2个冠状动脉树：一为LCA树，二是RCA树)；另外，在处理过程中也提取出少许其它血管，例如心脏静脉，及组织边缘，例如心脏房室，因此，需要进一步对血管树进行修剪。In the embodiment of the present invention, because the blood vessel curve fusion strategy is completed one by one on the basis of the initial N blood vessel curves, many overlapping curves will be caused when the number of blood vessel curves remains unchanged, which is the result of the same blood vessel path. Redundancy effect of multiple vascular centerlines. Similarly, the final vascular curve set S_N = {X, V, P, T}_N contains multiple vascular trees (the heart has only 2 coronary trees: one is the LCA tree, The second is the RCA tree); in addition, a few other blood vessels, such as cardiac veins, and tissue edges, such as cardiac atrioventricular, are also extracted during processing, so the vessel tree needs to be further pruned.

对初始化两条沿左右冠脉旅行的路径：分别从两条血管树根部的两个血管树的冠脉开口位置(起始点分别对应于x₀₁和x₀₂的位置)出发，向两个血管树末梢旅行，将遍历每一条血管中心线分支，对途经的各个血管分支按照以下准则进行分析：To initialize two paths that travel along the left and right coronary arteries: starting from the coronary opening positions of the two vascular trees at the root of the two vascular trees (the starting points correspond to the positions of x₀₁ and x₀₂ , respectively), to the two vascular trees. Peripheral travel will traverse each blood vessel centerline branch, and analyze each blood vessel branch along the way according to the following criteria:

获取新的血管曲线上逐点的局部测度值，将逐点的局部测度值构成测度值曲线，查找测度值曲线的波谷，截去波谷所在点的右侧曲线。Obtain the point-by-point local measurement value on the new blood vessel curve, form the point-by-point local measurement value into a measurement value curve, find the trough of the measurement value curve, and truncate the right curve of the point where the trough is located.

具体地，在血管增强图像

中测量每一条血管分支{X}_n上逐点的局部测度值，局部测度值计算公式为：Specifically, in the blood vessel enhanced image

Measure the point-by-point local measurement value of each blood vessel branch {X}_n in , and the calculation formula of the local measurement value is:

以上公式代表第n条血管曲线的第i点(x_i,i＝1,...,M_n)的局部测度值E_n(x_i)，采用相邻脊点区间

内强度的均值和方差计算，

为曲线长度。The above formula represents the local measurement value En (x_i ) of the i-_th point (x_i ,i=1,...,M_n ) of the n-th blood vessel curve, using the interval between adjacent ridge points

mean and variance calculation of inner strength,

is the length of the curve.

在新的血管曲线上，计算每一点相对于距离当前点为预设曲线长度的所在点的切线方向的变化程度，保留变化程度大于预设变化程度的血管曲线段。On the new blood vessel curve, the degree of change of each point in the tangential direction relative to the point where the distance from the current point is the preset curve length is calculated, and the blood vessel curve segments with the degree of change greater than the preset degree of change are retained.

具体地，在新的血管曲线上，在预设曲线长度L内，测量第i点相对于第i+L点的切线方向的变化程度，用预设变化程度

保留变化程度大于预设变化程度的血管曲线段，通常取

L＝30。Specifically, on the new blood vessel curve, within the preset curve length L, measure the degree of change in the tangential direction of the i-th point relative to the i+L-th point, and use the preset change degree

Retain the blood vessel curve segment with the change degree greater than the preset change degree, usually take the

L=30.

在本发明实施例中，将初始化的血管曲线与目标曲线进行融合，在目标曲线获取试探点，并根据夹角余弦阈值确定目标曲线的候选点，利用切向量夹角的最小值确定最终与目标曲线进行融合的融合点，构成新的血管曲线，提高了血管树提取的有效性及精确度，精确高效地提取血管分支和血管树是血管可视化、计算机辅助诊断、介入路径规划、手术导航的关键，对心血管介入手术中具有重要的临床意义。In the embodiment of the present invention, the initialized blood vessel curve and the target curve are fused, the tentative points are obtained on the target curve, and the candidate points of the target curve are determined according to the cosine threshold of the included angle, and the minimum value of the included angle of the tangent vector is used to determine the final value of the target curve. The fusion point of the curve fusion forms a new blood vessel curve, which improves the effectiveness and accuracy of the blood vessel tree extraction. Accurate and efficient extraction of blood vessel branches and blood vessel trees is the key to blood vessel visualization, computer-aided diagnosis, interventional path planning, and surgical navigation. , has important clinical significance in cardiovascular interventional surgery.

在本发明实施例中，血管树提取装置的各单元可由相应的硬件或软件单元实现，各单元可以为独立的软、硬件单元，也可以集成为一个软、硬件单元，在此不用以限制本发明。In this embodiment of the present invention, each unit of the apparatus for extracting the vascular tree may be implemented by a corresponding hardware or software unit, and each unit may be an independent software and hardware unit, or may be integrated into a software and hardware unit, which is not intended to limit the present invention. invention.

实施例四：Embodiment 4:

图5示出了本发明实施例四提供的血管树提取设备的结构示意图，为了便于说明，仅示出了与本发明实施例相关的部分。FIG. 5 shows a schematic structural diagram of a blood vessel tree extraction device provided in Embodiment 4 of the present invention. For convenience of description, only parts related to the embodiment of the present invention are shown.

本发明实施例的血管树提取设备5包括处理器50、存储器51以及存储在存储器51中并可在处理器50上运行的计算机程序52。该处理器50执行计算机程序52时实现上述各个血管树提取方法实施例中的步骤，例如图1所示的步骤101至104。或者，处理器50执行计算机程序52时实现上述各装置实施例中各模块/单元的功能，例如图4所示模块31至34的功能。The blood vesseltree extraction apparatus 5 of the embodiment of the present invention includes a processor 50 , amemory 51 , and acomputer program 52 stored in thememory 51 and executable on the processor 50 . When the processor 50 executes thecomputer program 52, the steps in each of the foregoing embodiments of the blood vessel tree extraction method are implemented, for example, steps 101 to 104 shown in FIG. 1 . Alternatively, when the processor 50 executes thecomputer program 52, the functions of the modules/units in the above device embodiments, for example, the functions of themodules 31 to 34 shown in FIG. 4 are implemented.

在本发明实施例中，将不连续的血管脊线作为初始化的血管曲线；在对血管曲线进行迭代的过程中，利用迭代后的血管曲线的端点，获取目标曲线的试探点，并根据夹角余弦阈值，从试探点中确定目标曲线的候选点，通过计算连接矢量与各候选点的切向量夹角的最小值，在候选点中确定目标曲线的融合点，该连接矢量为迭代后的血管曲线的端点与候选点的连接矢量；将当前血管曲线的端点与目标曲线中融合点的一侧曲线进行融合连接，构成新的血管曲线。In the embodiment of the present invention, the discontinuous blood vessel ridge line is used as the initialized blood vessel curve; in the process of iterating the blood vessel curve, the end point of the iterated blood vessel curve is used to obtain the test point of the target curve, and according to the included angle Cosine threshold, determine the candidate points of the target curve from the test points, and determine the fusion point of the target curve in the candidate points by calculating the minimum value of the angle between the connection vector and the tangent vector of each candidate point, and the connection vector is the iterative blood vessel The connection vector between the end point of the curve and the candidate point; the end point of the current blood vessel curve and the curve on one side of the fusion point in the target curve are fused and connected to form a new blood vessel curve.

在本发明实施例中，将初始化的血管曲线与目标曲线进行融合，在目标曲线获取试探点，并根据夹角余弦阈值确定目标曲线的候选点，利用切向量夹角的最小值确定最终与目标曲线进行融合的融合点，构成新的血管曲线，提高了血管树提取的有效性及精确度，精确高效地提取血管分支和血管树是血管可视化、计算机辅助诊断、介入路径规划、手术导航的关键，对心血管介入手术中具有重要的临床意义。In the embodiment of the present invention, the initialized blood vessel curve and the target curve are fused, the tentative points are obtained on the target curve, and the candidate points of the target curve are determined according to the cosine threshold of the included angle, and the minimum value of the included angle of the tangent vector is used to determine the final value of the target curve. The fusion point of the curve fusion forms a new blood vessel curve, which improves the effectiveness and accuracy of the blood vessel tree extraction. Accurate and efficient extraction of blood vessel branches and blood vessel trees is the key to blood vessel visualization, computer-aided diagnosis, interventional path planning, and surgical navigation. , has important clinical significance in cardiovascular interventional surgery.

实施例五：Embodiment 5:

在本发明实施例中，提供了一种计算机可读存储介质，该计算机可读存储介质存储有计算机程序，该计算机程序被处理器执行时实现上述各个血管树提取方法实施例中的步骤，例如，图1所示的步骤101至104。或者，该计算机程序被处理器执行时实现上述各装置实施例中各模块/单元的功能，例如图4所示模块31至34的功能。In an embodiment of the present invention, a computer-readable storage medium is provided, and the computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the steps in each of the foregoing embodiments of the blood vessel tree extraction method are implemented, for example, , steps 101 to 104 shown in FIG. 1 . Alternatively, when the computer program is executed by the processor, the functions of the modules/units in the above device embodiments, for example, the functions of themodules 31 to 34 shown in FIG. 4 , are implemented.

将不连续的血管脊线作为初始化的血管曲线；在对血管曲线进行迭代的过程中，利用迭代后的血管曲线的端点，获取目标曲线的试探点，并根据夹角余弦阈值，从试探点中确定目标曲线的候选点，通过计算连接矢量与各候选点的切向量夹角的最小值，在候选点中确定目标曲线的融合点，该连接矢量为迭代后的血管曲线的端点与候选点的连接矢量；将当前血管曲线的端点与目标曲线中融合点的一侧曲线进行融合连接，构成新的血管曲线。The discontinuous blood vessel ridge line is used as the initialized blood vessel curve; in the process of iterating the blood vessel curve, the end point of the iterative blood vessel curve is used to obtain the test point of the target curve, and according to the cosine threshold of the included angle, the test point is obtained from the test point. Determine the candidate points of the target curve, and determine the fusion point of the target curve in the candidate points by calculating the minimum value of the angle between the connection vector and the tangent vector of each candidate point. The connection vector is the endpoint of the iterative blood vessel curve and the candidate point. Connection vector; the endpoint of the current blood vessel curve is fused and connected to the curve on one side of the fusion point in the target curve to form a new blood vessel curve.

本发明实施例的计算机可读存储介质可以包括能够携带计算机程序代码的任何实体或装置、记录介质，例如，ROM/RAM、磁盘、光盘、闪存等存储器。The computer-readable storage medium of the embodiments of the present invention may include any entity or device capable of carrying computer program codes, recording medium, for example, memory such as ROM/RAM, magnetic disk, optical disk, flash memory, and the like.

以上所述仅为本发明的较佳实施例而已，并不用以限制本发明，凡在本发明的精神和原则之内所作的任何修改、等同替换和改进等，均应包含在本发明的保护范围之内。The above descriptions are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present invention shall be included in the protection of the present invention. within the range.

Claims (10)

1. A vessel tree extraction method is characterized by comprising the following steps:

taking discontinuous blood vessel ridge lines as initialized blood vessel curves;

in the process of iterating the blood vessel curve, utilizing the endpoint of the iterated blood vessel curve to obtain a tentative point of a target curve, calculating a cosine value of a direction included angle between a direction vector of the endpoint and a connecting direction vector from the endpoint to the tentative point, and determining a current tentative point as a candidate point of the target curve when the cosine value meets a cosine threshold of the included angle;

calculating an included angle between a connecting vector and tangent vectors of the candidate points, and determining a corresponding candidate point with the smallest included angle as a fusion point of the target curve, wherein the connecting vector is a connecting vector between an end point of the iterated vessel curve and the candidate point;

and carrying out fusion connection on the end point of the current blood vessel curve and the curve on one side of the fusion point in the target curve to form a new blood vessel curve.

2. The method of claim 1, wherein the step of using the discontinuous vessel ridge as the initialized vessel curve comprises:

and acquiring position coordinates and tangent vectors of all points in the blood vessel curve, and marking end point attributes and curve attributes of the blood vessel curve.

3. The method of claim 1, wherein the step of obtaining tentative points of a target curve using the end points of the iterated vessel curve during the iteration of the vessel curve and determining candidate points of the target curve from the tentative points according to an angle cosine threshold comprises:

acquiring the tentative point in a preset range of the endpoint of the iterated blood vessel curve, and calculating a cosine value of a direction included angle between a direction vector of the endpoint and a connecting direction vector from the endpoint to the tentative point;

and when the cosine value meets the cosine threshold of the included angle, determining the current probing point as a candidate point of the target curve.

4. The method of claim 1, wherein the step of fusion-connecting the end points of the current vessel curve with the curve on one side of the fusion point in the target curve to form a new vessel curve comprises:

performing fusion connection on the left end point of the blood vessel curve and the left curve of the fusion point in the target curve; or

Performing fusion connection on the left end point of the blood vessel curve and the right side curve of the fusion point in the target curve; or

Carrying out fusion connection on the right end point of the blood vessel curve and the left side curve of the fusion point in the target curve; or

And carrying out fusion connection on the right end point of the blood vessel curve and the right curve of the fusion point in the target curve.

5. The method of claim 1, wherein the step of fusion-connecting the end points of the current vessel curve with the curve on one side of the fusion point in the target curve to form a new vessel curve comprises:

and acquiring point-by-point local measurement values on the new blood vessel curve, forming the point-by-point local measurement values into a measurement value curve, searching a trough of the measurement value curve, and cutting a right curve of the point where the trough is located.

6. The method of claim 1, wherein the step of fusion-connecting the end points of the current vessel curve with the curve on one side of the fusion point in the target curve to form a new vessel curve further comprises:

calculating the change degree of each point relative to the tangential direction of the point which is at the preset curve length from the current point on the new blood vessel curve;

and reserving the vessel curve segment with the change degree larger than the preset change degree.

7. A vessel tree extraction device, characterized in that the device comprises:

the initialization unit is used for taking discontinuous blood vessel ridge lines as initialized blood vessel curves;

the candidate point determining unit is used for acquiring a tentative point of a target curve by using an endpoint of the iterated blood vessel curve in the process of iterating the blood vessel curve, calculating a cosine value of a direction included angle between a direction vector of the endpoint and a connecting direction vector from the endpoint to the tentative point, and determining a current tentative point as a candidate point of the target curve when the cosine value meets a cosine threshold of the included angle;

a fusion point determining unit, configured to calculate an included angle between a connection vector and a tangent vector of each candidate point, and determine a corresponding candidate point with a smallest included angle as a fusion point of the target curve, where the connection vector is a connection vector between an end point of the iterated vascular curve and the candidate point; and

and the fusion connecting unit is used for performing fusion connection on the end point of the current blood vessel curve and the curve on one side of the fusion point in the target curve to form a new blood vessel curve.

8. The apparatus of claim 7, wherein the apparatus further comprises:

and the curve cutting unit is used for acquiring point-by-point local measurement values on the new blood vessel curve, forming the point-by-point local measurement values into a measurement value curve, searching a trough of the measurement value curve and cutting a right curve of the point where the trough is located.

9. A vessel tree extraction device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor implements the steps of the method according to any of claims 1 to 6 when executing the computer program.

10. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 6.

Images