CN114299057A - Extraction method and storage medium of blood vessel centerline - Google Patents

Abstract

The application relates to a method, a device, computer equipment, a storage medium and a computer program product for extracting a blood vessel center line, wherein a first blood vessel segmentation image comprising at least one blood vessel segment is obtained by acquiring an initial black blood image of an object to be detected and inputting the initial black blood image into a preset first segmentation model; extracting the center line of each blood vessel segment in the first blood vessel segment image to obtain a blood vessel center line image corresponding to the initial black blood image; the automatic extraction mode that the center line of the blood vessel can be directly obtained from the black blood image can be realized, the bright blood image corresponding to the black blood image is not needed, the bright blood image is not needed to be additionally obtained, and the quality problem and the registration accuracy problem of the bright blood image are not needed to be considered; after the initial black blood image is obtained, the method in the embodiment can be used for directly identifying and processing the initial black blood image to obtain the blood vessel centerline image corresponding to the initial black blood image, and the extraction efficiency is high.

Description

Extraction method and storage medium of blood vessel centerline

technical field

The present application relates to the technical field of computer and image processing, and in particular, to a method, apparatus, computer equipment, storage medium and computer program product for extracting a blood vessel centerline.

Background technique

In the diagnosis of cerebrovascular disease, it is usually necessary to identify the main cerebral blood vessels in the image according to the image obtained by magnetic resonance, and then, the diseased area of the cerebrovascular can be detected based on the identified cerebral blood vessels to obtain a diagnosis result. For a black blood image of a cerebral blood vessel, when identifying a cerebral blood vessel in the black blood image, it is usually necessary to first determine the centerline of the cerebral blood vessel in the black blood image, and then extract the black blood image based on the centerline of the cerebral blood vessel. brain blood vessels.

In the traditional technology, the process of automatically identifying the blood vessel centerline of the black blood image is as follows: manually or automatically obtaining the blood vessel centerline in the bright blood image corresponding to the black blood image, and then, according to the registration algorithm, the blood vessels in the bright blood image are The center line is registered to the black blood image, and then the registered blood vessel center line is fine-tuned to obtain the blood vessel center line of the black blood image.

However, in the existing method for obtaining the blood vessel centerline of the black blood image, when the quality of the bright blood image is not high or the registration algorithm is inaccurate, the accuracy of the obtained blood vessel centerline of the black blood image is likely to be poor.

SUMMARY OF THE INVENTION

Based on this, it is necessary to provide a method, device, computer equipment, and computer-readable storage for extracting the blood vessel centerline, which can extract the blood vessel centerline only through the black blood image, thereby improving the extraction accuracy of the blood vessel centerline. Media and Computer Program Products.

In a first aspect, the present application provides a method for extracting a blood vessel centerline. The method includes:

Obtain the initial black blood image of the object to be tested;

Inputting the initial black blood image into a preset first segmentation model to obtain a first blood vessel segmented image; wherein the first blood vessel segmented image includes at least one blood vessel segment;

The center line of each blood vessel segment in the first blood vessel segment image is extracted to obtain a blood vessel center line image corresponding to the initial black blood image.

In one embodiment, the centerline of each blood vessel segment in the first blood vessel segment image is extracted to obtain a blood vessel centerline image corresponding to the initial black blood image, including:

For each vessel segment, extract the centerline of each vessel segment;

Based on the initial black blood image and the preset blood vessel connection strategy, the centerline of each blood vessel segment is connected to obtain the blood vessel centerline image corresponding to the initial black blood image.

In one embodiment, extracting the centerline of each blood vessel segment includes:

For a blood vessel segment, determining a first endpoint and a second endpoint of the vessel segment;

Based on the first blood vessel segment image, the first end point and the second end point of the blood vessel segment are connected to obtain the center line of the blood vessel segment.

In one embodiment, determining the first endpoint and the second endpoint of the vessel segment includes:

The blood vessel segment includes multiple sub-vessel segments;

determining the first endpoint and the second endpoint of each sub-vessel segment;

Correspondingly, based on the first blood vessel segment image, connect the first end point and the second end point of the blood vessel segment to obtain the center line of the blood vessel segment, including:

For each sub-vessel segment, based on the first blood vessel segment image, connect the first endpoint and the second endpoint of the sub-vessel segment to obtain the centerline of the sub-vessel segment;

Based on the initial black blood image, connect the centerlines of each sub-vessel segment to obtain the centerline of the vessel segment.

In one embodiment, the method further includes:

determining at least one blood vessel bifurcation point based on the blood vessel centerline image and the first blood vessel segment image;

Based on each blood vessel bifurcation point, each blood vessel centerline in the blood vessel centerline image is segmented to obtain a segmented image of the blood vessel centerline.

In one embodiment, determining at least one blood vessel bifurcation point based on the blood vessel centerline image and the first blood vessel segment image includes:

determining the position of the first bifurcation point in the first blood vessel segmented image, and the position of the second bifurcation point in the blood vessel centerline image;

Determine whether the position of the second bifurcation point is within the preset range of the position of the first bifurcation point;

When the position of the second bifurcation point is within a preset range of the position of the first bifurcation point, the second bifurcation point is determined as a blood vessel bifurcation point.

In one of the embodiments, the method further includes: in the case that the position of the second bifurcation point is not within the preset range of the position of the first bifurcation point, returning to re-execute the strategy based on the initial black blood image and the blood vessel connection, Connect the centerlines of each blood vessel segment to obtain an adjusted blood vessel centerline image, and perform the steps of determining the position of the second bifurcation point in the adjusted blood vessel centerline image, until the position of the second bifurcation point is located at within the preset range of the position of the first bifurcation point.

In one embodiment, the method further includes:

Based on the initial black blood image, obtain a cross-sectional image corresponding to each point on the centerline of each blood vessel in the blood vessel centerline image;

Each cross-sectional image is processed to obtain a target blood vessel image corresponding to the blood vessel centerline image.

In one embodiment, each cross-sectional image is processed to obtain a target blood vessel image corresponding to the blood vessel centerline image, including:

Inputting each cross-sectional image into the second segmentation model, respectively, to obtain a segmentation result image corresponding to each cross-sectional image;

Based on the segmentation result image corresponding to each cross-sectional image, determine whether the blood vessel structure in each cross-sectional image satisfies a preset blood vessel structure rule;

When the blood vessel structure in the cross-sectional image does not meet the preset blood vessel structure rule, the cross-sectional image that does not satisfy the blood vessel structure rule is removed, and the target blood vessel image is obtained according to the remaining cross-sectional images that satisfy the blood vessel structure rule.

In one embodiment, the method further includes:

Based on each blood vessel bifurcation point, each blood vessel in the target blood vessel image is segmented to obtain a target blood vessel segmented image corresponding to the target blood vessel image.

In one embodiment, the method further includes:

Based on the blood vessel centerline image, detect whether the target tissue exists in the blood vessel in the initial black blood image;

When the target tissue exists in the blood vessel in the initial black blood image, the target tissue is segmented based on the initial black blood image to obtain a segmentation result of the target tissue.

In a second aspect, the present application also provides a device for extracting a blood vessel centerline. The device includes:

The first acquisition module is used to acquire the initial black blood image of the object to be tested;

The second acquisition module is configured to input the initial black blood image into a preset first segmentation model to obtain a first blood vessel segmented image; wherein, the first blood vessel segmented image includes at least one blood vessel segment;

The third acquisition module is configured to extract the centerline of each blood vessel segment in the first blood vessel segmented image, and obtain the blood vessel centerline image corresponding to the initial black blood image.

In a third aspect, the present application also provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

Obtain the initial black blood image of the object to be tested;

Inputting the initial black blood image into a preset first segmentation model to obtain a first blood vessel segmented image; wherein the first blood vessel segmented image includes at least one blood vessel segment;

The center line of each blood vessel segment in the first blood vessel segment image is extracted to obtain a blood vessel center line image corresponding to the initial black blood image.

In a fourth aspect, the present application also provides a computer-readable storage medium. The computer-readable storage medium has a computer program stored thereon, and when the computer program is executed by the processor, the following steps are implemented:

Obtain the initial black blood image of the object to be tested;

Inputting the initial black blood image into a preset first segmentation model to obtain a first blood vessel segmented image; wherein the first blood vessel segmented image includes at least one blood vessel segment;

The center line of each blood vessel segment in the first blood vessel segment image is extracted to obtain a blood vessel center line image corresponding to the initial black blood image.

In a fifth aspect, the present application also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, implements the following steps:

Obtain the initial black blood image of the object to be tested;

Inputting the initial black blood image into a preset first segmentation model to obtain a first blood vessel segmented image; wherein the first blood vessel segmented image includes at least one blood vessel segment;

The center line of each blood vessel segment in the first blood vessel segment image is extracted to obtain a blood vessel center line image corresponding to the initial black blood image.

The above-mentioned method, device, computer equipment, storage medium and computer program product for extracting the centerline of blood vessels, the computer equipment obtains the initial black blood image of the object to be measured by acquiring the initial black blood image and inputting the initial black blood image into the preset first segmentation model to obtain A first blood vessel segment image including at least one blood vessel segment, and then extracting the centerline of each blood vessel segment in the first blood vessel segment image to obtain a blood vessel centerline image corresponding to the initial black blood image; The automatic extraction method of the blood vessel centerline can be directly obtained from the black blood image, no need to use the bright blood image corresponding to the black blood image, and no need to additionally obtain the bright blood image. Of course, there is no need to consider the quality of the bright blood image. The problem and the accuracy of registration; after the initial black blood image is obtained, the method in this embodiment can directly identify and process the initial black blood image to obtain the blood vessel center line corresponding to the initial black blood image. Image, the operation is simple and quick, time-saving and labor-saving, the obtained blood vessel centerline image has high accuracy and high extraction efficiency.

Description of drawings

1 is a schematic flowchart of a method for extracting a blood vessel centerline in one embodiment;

2 is a schematic structural diagram of a first blood vessel segmented image in one embodiment;

3 is a schematic structural diagram of a blood vessel centerline image in one embodiment;

4 is a schematic flowchart of a method for extracting a blood vessel centerline in another embodiment;

5 is a schematic flowchart of a method for extracting a blood vessel centerline in another embodiment;

6 is a schematic flowchart of a method for extracting a blood vessel centerline in another embodiment;

7 is a schematic diagram of a centerline extraction process of blood vessel segmentation in one embodiment;

8 is a schematic flowchart of a method for extracting a blood vessel centerline in another embodiment;

9 is a schematic flowchart of a method for extracting a blood vessel centerline in another embodiment;

10 is a schematic flowchart of a method for extracting a blood vessel centerline in another embodiment;

11 is a schematic flowchart of a method for extracting a blood vessel centerline in another embodiment;

12 is a complete schematic diagram of a change process of a method for extracting a blood vessel centerline in one embodiment;

13 is a structural block diagram of an apparatus for extracting blood vessel centerline in one embodiment;

Figure 14 is a diagram of the internal structure of a computer device in one embodiment.

Detailed ways

In order to make the purpose, technical solutions and advantages of the present application more clearly understood, the present application will be described in further 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 application, but not to limit the present application.

The method for extracting the blood vessel centerline provided by the embodiments of the present application can be applied to computer equipment. The computer equipment can be a terminal or a server, and the terminal can be, but is not limited to, various personal computers, notebook computers, smart phones, Tablet computers, scanning devices, and processing devices connected to the scanning devices, etc., the server can be implemented by an independent server or a server cluster composed of multiple servers.

In one embodiment, as shown in FIG. 1, a method for extracting a blood vessel centerline is provided, and the method is applied to the above-mentioned computer equipment as an example to illustrate, including the following steps:

Step 101: Obtain an initial black blood image of the object to be tested.

Optionally, the initial black blood image may be collected from a collection device, may also be obtained from a server, or may be a historical initial black blood image obtained locally from the device, etc.; The acquisition method of the initial black blood image is not limited. In addition, the initial black blood image may be a black blood sequence of different forms, including but not limited to T1 enhanced image, T1 image, T2 image, proton density image, etc. The form of the initial black blood image is not limited in this embodiment.

Step 102: Input the initial black blood image into a preset first segmentation model to obtain a first segmented image of blood vessels.

The first blood vessel segment image includes at least one blood vessel segment, and each blood vessel segment may be a part of blood vessels distributed in different tissues or organs, or may be different types of blood vessels in the same tissue or organ, for example: arteries Blood vessels, veins, capillaries, etc. Optionally, in the first blood vessel segmented image, different blood vessel segments can be marked by different line types, for example, different labeling can be performed according to line color, line thickness, or line shape. Differential labeling, etc.; identification information can also be used to distinguish different blood vessel segments, for example: text identification, digital identification, symbol identification, etc.; Segmentation, this embodiment does not limit the labeling manner of the blood vessel segment.

Optionally, the first segmentation model may be a segmentation model obtained after training the first initial segmentation network by using a plurality of black blood sample images and a blood vessel segment label image corresponding to each black blood sample image. The initial segmentation network may be an existing deep learning network of any type, or may be a segmentation network combining various types of networks, etc., which is not limited in this embodiment.

Through the first segmentation model, the blood vessel segmentation process can be performed on the initial black blood image to obtain a first blood vessel segmented image corresponding to the initial black blood image; optionally, the number of blood vessel segments in the first blood vessel segmented image It can be the same as the number of blood vessel segments in the initial black blood image, and the effect of blood vessel segmentation is better. For example, as shown in FIG. 2 , it is the first segmented image of blood vessels after blood vessel segmentation is performed on the initial black blood image.

Step 103 , extracting the centerline of each blood vessel segment in the first blood vessel segment image to obtain a blood vessel centerline image corresponding to the initial black blood image.

Optionally, after acquiring the first blood vessel image including at least one blood vessel segment corresponding to the initial black blood image, centerline extraction can be performed on each blood vessel segment in the first blood vessel segment image, to obtain each blood vessel segment. The centerline corresponding to the blood vessel segment can be obtained to obtain the blood vessel centerline image corresponding to the initial black blood image, as shown in Figure 3.

Optionally, a preset centerline extraction algorithm can be used to extract the centerline of each blood vessel segment in the first blood vessel segmented image, or an online learning algorithm or the like can be used to extract each blood vessel segment in the first blood vessel segmented image. The centerline of a blood vessel segment, wherein the centerline extraction algorithm may be a model-based centerline extraction algorithm, or a non-model-based centerline extraction algorithm. The method for extracting the blood vessel centerline in this embodiment and the centerline The principle and method of the extraction algorithm are not limited.

In the above method for extracting the blood vessel centerline, the computer equipment obtains the first blood vessel segment including at least one blood vessel segment by acquiring the initial black blood image of the object to be measured and inputting the initial black blood image into the preset first segmentation model. Then, the center line of each blood vessel segment in the first blood vessel segment image is extracted to obtain the blood vessel center line image corresponding to the initial black blood image; it is realized that the blood vessel center can be directly obtained only from the black blood image. The automatic extraction method of the line does not need to use the bright blood image corresponding to the black blood image, nor does it need to obtain additional bright blood images. Of course, there is no need to consider the quality of the bright blood images and the accuracy of registration; After reaching the initial black blood image, the method in this embodiment can directly obtain the blood vessel centerline image corresponding to the initial black blood image by identifying and processing the initial black blood image. The operation is simple and fast, saving time and effort, so The obtained blood vessel centerline images have high accuracy and high extraction efficiency.

FIG. 4 is a schematic flowchart of a method for extracting a blood vessel centerline in another embodiment. This embodiment relates to an optional implementation process in which a computer device extracts the centerline of each blood vessel segment in the first blood vessel segmented image to obtain a blood vessel centerline image corresponding to the initial black blood image; as shown in FIG. 4 As shown, on the basis of the foregoing embodiment, the foregoing step 103 includes:

Step 401, for each blood vessel segment, extract the center line of each blood vessel segment.

Regarding the implementation manner of extracting the centerline of the blood vessel segment, reference may be made to the relevant content in the foregoing step 103, which will not be repeated here. Optionally, in this embodiment, different line colors may be used to mark different blood vessel segments.

Step 402 , based on the initial black blood image and a preset blood vessel connection strategy, connect the centerlines of each blood vessel segment to obtain a blood vessel centerline image corresponding to the initial black blood image.

The preset blood vessel connection strategy may include connection rules for multiple blood vessels, and the connection rules for each blood vessel may include multiple connection rules for blood vessel segments located in different tissues or organs; for example, for a complete blood vessel Arterial blood vessels, which can shuttle through multiple different tissues or organs, can use different line colors to distinguish the multiple blood vessel segments of the arterial blood vessels in different tissues or organs, and can preset the corresponding arterial blood vessels. The order of connections between the vessel segments.

Optionally, after extracting the centerline of each blood vessel segment, the computer device can connect the centerlines of the blood vessel segments with associated relationships in each blood vessel segment based on the initial black blood image and the blood vessel connection strategy. , obtain a blood vessel centerline image corresponding to the initial black blood image, wherein the blood vessel centerline image includes at least one complete blood vessel.

Optionally, the computer device may determine, based on the blood vessel connection strategy and the blood vessel connection sequence of each blood vessel in the blood vessel connection strategy, two blood vessel segments to be connected that have an associated relationship in the first blood vessel currently to be connected, Then, based on the initial black blood image, the optimal path algorithm can be used to connect the center lines of the two blood vessel segments to be connected, and so on, until the center of each blood vessel segment in the first blood vessel Until the line is connected, the first blood vessel center line corresponding to the first blood vessel is obtained; similarly, after the center line of the first blood vessel is connected, the center line of the second blood vessel can be connected, and so on, until there is a The centerline image of the blood vessel can be obtained until the centerlines of all the blood vessel segments in the associated relationship are connected.

In this embodiment, for each blood vessel segment, the computer equipment extracts the center line of each blood vessel segment, and connects the center line of each blood vessel segment based on the initial black blood image and the preset blood vessel connection strategy, to obtain The blood vessel centerline image corresponding to the initial black blood image can improve the extraction accuracy of the blood vessel centerline, and can improve the efficiency of the blood vessel centerline improvement.

FIG. 5 is a schematic flowchart of a method for extracting a blood vessel centerline in another embodiment. This embodiment relates to an optional implementation process for extracting the centerline of each blood vessel segment by a computer device; as shown in FIG. 5 , on the basis of the foregoing embodiment, the foregoing step 401 includes:

Step 501, for the blood vessel segment, determine the first endpoint and the second endpoint of the blood vessel segment.

The first end point of the blood vessel segment may be the center point of the head end of the blood vessel segment, and the second end point of the blood vessel segment may be the center point of the tail end of the blood vessel segment.

Step 502: Based on the first blood vessel segment image, connect the first end point and the second end point of the blood vessel segment to obtain the center line of the blood vessel segment.

Optionally, when determining the center line of the blood vessel segment in the first blood vessel segmented image, after determining the first endpoint and the second endpoint of the blood vessel segment, it can be based on the first blood vessel segmented image. The segmented blood vessel connects the first end point and the second end point of the blood vessel segment to form the centerline of the blood vessel segment; optionally, based on the segmented blood vessel in the first blood vessel segmented image, a The vessel thinning algorithm obtains the center points at multiple positions of the segmented vessel, and connects the center points at the multiple positions with the first and second endpoints of the vessel segment to obtain the vessel segment. The center line of the segment; or based on the segmented blood vessel in the first blood vessel segmented image, the optimal path algorithm can be used to connect the first end point and the second end point of the blood vessel segment to obtain the center of the blood vessel segment. It should be noted that, in this embodiment, the specific implementation of connecting the first end point and the second end point of the blood vessel segment to form the center line of the blood vessel segment is not limited. There are any kind of centerline determination methods to achieve.

In this embodiment, for the blood vessel segment, the computer device connects the first end point and the second end point of the blood vessel segment by determining the first end point and the second end point of the blood vessel segment and based on the first blood vessel segment image The end point is obtained to obtain the center line of the blood vessel segment, the obtained center line of the blood vessel segment has high accuracy, the realization process is simple, and the extraction efficiency of the center line is high.

In an optional embodiment of the present application, based on the influence of some objective factors, the different blood vessel segments in the initial black blood image of the object to be tested may have problems of occlusion and poor developing effect, resulting in the first segmentation Each blood vessel segment in the first blood vessel segment image corresponding to the initial black blood image output by the model may be broken, that is, for the same blood vessel segment, there may be multiple sub-vessel segments (for example: in the There may be multiple blood vessel segments of the same color in the first blood vessel segment image, and the multiple blood vessel segments of the same color are part of the blood vessels in the same tissue or organ); When the centerline of the segment is determined, the centerlines of the multiple sub-vessel segments corresponding to the vessel segment can be determined respectively, and then the centerlines of the respective sub-vessel segments are connected to obtain the centerline of the vessel segment.

FIG. 6 is a schematic flowchart of a method for extracting a blood vessel centerline in another embodiment. This embodiment relates to an optional implementation process for a computer device to extract the centerline of each blood vessel segment when the blood vessel segment includes multiple sub-vessel segments; as shown in FIG. 6 , in the above implementation process On the basis of the example, the above step 501 includes:

Step 601: Determine the first endpoint and the second endpoint of each sub-vessel segment.

The first end point of the sub-vessel segment may be the center point of the head end of the sub-vessel segment, and the second end point of the sub-vessel segment may be the center point of the tail end of the sub-vessel segment.

Correspondingly, the above step 502 includes:

Step 602 , for each sub-vessel segment, based on the first blood vessel segment image, connect the first end point and the second end point of the sub-vessel segment to obtain the center line of the sub-vessel segment.

The manner of determining the centerline of the sub-vessel segment here may refer to the implementation manner of determining the centerline of the vessel segment in the foregoing step 502, and details are not described herein again.

Step 603 , based on the initial black blood image, connect the centerlines of each sub-vessel segment to obtain the centerline of the blood vessel segment.

Optionally, after determining the centerlines of multiple sub-vessel segments of a certain vessel segment, the connection sequence of the centerlines of each sub-vessel segment can be determined based on the initial black blood image, and then, according to the connection Sequentially, connect the centerlines of two adjacent sub-vessel segments to obtain the centerline of the blood vessel segment; optionally, when connecting the centerlines of two adjacent sub-vessel segments, for these two sub-vessel segments For the vacant part between the center lines of the two sub-vessel segments, the optimal path algorithm can be used, and with the help of the initial black blood image, a partial center line between the center lines of the two sub-vessel segments can be generated, so as to connect the two sub-vessel segments. Of course, other connection methods or connection algorithms may also be used to connect the centerlines of two adjacent sub-vessel segments, which are not limited in this embodiment of the present application.

In this embodiment, when the blood vessel segment includes multiple sub-vessel segments, the first endpoint and the second endpoint of each sub-vessel segment can be determined, and for each sub-vessel segment, based on the first blood vessel segment segment image, connect the first end point and the second end point of the sub-vessel segment to obtain the center line of the sub-vessel segment; then, based on the initial black blood image, connect the center lines of each sub-vessel segment to obtain the blood vessel segment The centerline of the obtained blood vessel segment has high accuracy, and the obtained centerline of the blood vessel segment is more complete, which improves the accuracy and completeness of the centerline extraction.

In another optional embodiment of the present application, when the blood vessel segment includes multiple sub-vessel segments, the two longest sub-vessel segments may also be determined from the multiple sub-vessel segments corresponding to the blood vessel segment segment, and based on the two longest sub-vessel segments, the centerline of the vessel segment is obtained; optionally, for a vessel segment (as shown in Figure 7(a), the Structure diagram), the longest sub-vessel segment and the second-longest sub-vessel segment can be determined from the multiple sub-vessel segments of the vessel segment, and the first endpoint of the longest sub-vessel segment and the second-longest sub-vessel segment are determined. The first endpoint is the first endpoint of the vessel segment, and the second endpoint of the longest sub-vessel segment and the second endpoint of the second-longest sub-vessel segment are determined to be the second endpoint of the vessel segment; that is, Determine the first endpoint and the second endpoint of the longest sub-vessel segment, and determine the first endpoint and the second endpoint of the second-longest sub-vessel segment; then, for the longest sub-vessel segment, connect the longest sub-vessel the first endpoint and the second endpoint of the segment to obtain the first centerline of the longest sub-vessel segment, and for the second-longest sub-vessel segment, connecting the first and second endpoints of the second-longest sub-vessel segment, Obtain the second centerline of the sub-longest sub-vessel segment (as shown in Figure 7(b), which is a schematic structural diagram of the centerline of the longest sub-vessel segment and the center line of the sub-longest sub-vessel segment in the blood vessel segment); then , the first centerline of the longest sub-vessel segment and the second centerline of the second-longest sub-vessel segment can be connected to represent the centerline of the vessel segment (as shown in Figure 7(c), for the blood vessel Schematic diagram of the segmented centerline).

Optionally, when connecting the first end point and the second end point of the longest sub-vessel segment, an optimal path algorithm can be used to connect the first end of the longest sub-vessel segment based on the above-mentioned first blood vessel segment image. point and the second endpoint of the longest sub-vessel segment to obtain the centerline of the longest sub-vessel segment; similarly, connect the first endpoint of the second-longest sub-vessel segment and the second endpoint of the second-long sub-vessel segment , based on the above-mentioned first blood vessel segment image, the optimal path algorithm can be used to connect the first end point of the second longest sub-vessel segment and the second end point of the sub-long sub-vessel segment to obtain the second end of the sub-long sub-vessel Centerline; then, based on the above-mentioned initial black blood image, an optimal path algorithm can be used to connect the centerline of the longest sub-vessel segment and the centerline of the second-longest sub-vessel segment to obtain the centerline of the vessel segment.

In this embodiment, when a blood vessel segment includes multiple sub-vessel segments, the blood vessel may be determined based on the longest sub-vessel segment and the second-longest sub-vessel segment among the multiple sub-vessel segments corresponding to the blood vessel segment The centerline of the segment, that is, the vessel segment is represented by the two longest sub-vessel segments in the vessel segment, which can improve the extraction rate of the centerline of the vessel segment, and at the same time ensure the vessel segment. accuracy of the centerline.

FIG. 8 is a schematic flowchart of a method for extracting a blood vessel centerline in another embodiment. This embodiment relates to an optional implementation process in which the computer device further determines the segmented image of the blood vessel centerline on the basis of the above-mentioned blood vessel centerline image; as shown in FIG. 8 , on the basis of the above embodiment, The above method also includes:

Step 801: Determine at least one blood vessel bifurcation point based on the blood vessel centerline image and the first blood vessel segmented image.

Wherein, the first blood vessel segment image includes a plurality of blood vessel segments, the plurality of blood vessel segments include a plurality of blood vessels of different types, and each blood vessel includes at least one blood vessel segment located in a different tissue or organ. In the above-mentioned embodiment, by acquiring the centerline of each blood vessel segment and connecting multiple blood vessel segments with associated relationships that belong to the same blood vessel, a complete blood vessel centerline including multiple blood vessels is obtained, that is, the blood vessel. Centerline image; after the blood vessel centerline image is obtained, each complete blood vessel in the blood vessel centerline may be segmented to obtain a blood vessel centerline segmented image corresponding to the first blood vessel segmented image. Each vessel segment centerline in the vessel centerline segmented image corresponds one-to-one with each vessel segment in the first vessel segment image, but compared to the image based on the initial black blood image and the first vessel segment image For the obtained centerline of each blood vessel segment, the centerline of each blood vessel segment obtained based on the blood vessel centerline image and the first blood vessel segment image will be more complete, and the two blood vessel segments are connected to each other. The connectivity between the centerlines of the blood vessels will also be better, while ensuring the integrity of each blood vessel, it can also accurately distinguish the distribution of the blood vessel segments of each blood vessel in different tissues or organs.

Optionally, before segmenting the blood vessel centerline image, it is necessary to determine a plurality of bifurcation points, and the plurality of blood vessel bifurcation points are connection points between each blood vessel segment, so that the computer equipment can The bifurcation point performs segmentation processing on a plurality of blood vessels in the blood vessel centerline image; when determining the bifurcation point, the plurality of bifurcation points in the first blood vessel segmented image can be determined, and the Then, by comparing the plurality of bifurcation points in the first blood vessel segmented image and the plurality of bifurcation points in the blood vessel centerline image, the final segmentation of the blood vessel centerline image can be obtained. segment of multiple vascular bifurcation points.

Optionally, the position of the first bifurcation point in the first blood vessel segmented image and the position of the second bifurcation point in the blood vessel centerline image may be determined, wherein the first bifurcation point and the second bifurcation point are determined. The bifurcation point is the bifurcation point between the same blood vessels in the first blood vessel segmented image and the blood vessel centerline image; then, it can be determined whether the position of the second bifurcation point is in the preset position of the first bifurcation point In the case where it is determined that the position of the second bifurcation point is within the preset range of the position of the first bifurcation point, the second bifurcation point can be determined as the blood vessel bifurcation point, that is, in this case, the description The cross-connection between several blood vessels in the first blood vessel segmented image is basically consistent with the cross-connection between these blood vessels in the blood vessel centerline, which also shows that the centerline of each blood vessel in the blood vessel centerline image is the same as the initial one. The matching degree of the actual blood vessels in the black blood image is higher, and the obtained blood vessel centerline is more accurate. Therefore, in this case, the second bifurcation point determined from the blood vessel centerline image can be directly used as the final pair of blood vessel centers. Line images of vessel bifurcation points for segmentation.

Optionally, in the case that the position of the second bifurcation point is not within the preset range of the position of the first bifurcation point, it is explained that the result obtained after connecting the centerlines of the multiple blood vessel segments with the associated relationship is described. There is a difference between the center line of the complete blood vessel and the actual blood vessel in the initial black blood image; at this time, you can return to re-execute the initial black blood image and the blood vessel connection strategy to connect the center line of each blood vessel segment to obtain the adjusted blood vessel center Then, the step of determining the position of the second bifurcation point in the adjusted blood vessel centerline image may be continued, and it is determined whether the position of the second bifurcation point in the adjusted blood vessel centerline image is within the first If the position of the first bifurcation point in the blood vessel segmented image is within the preset range, if it is within the preset range, the second bifurcation point in the adjusted blood vessel centerline image may be used as the final segment of the blood vessel centerline image. The blood vessel bifurcation point; if it is still not there, you can go back and continue to execute the centerline of each blood vessel segment based on the initial black blood image and the blood vessel connection strategy to obtain the adjusted blood vessel centerline image, and the subsequent steps until the first Until the position of the bifurcation point is within the preset range of the position of the first bifurcation point.

Step 802: Perform segmentation processing on each blood vessel centerline in the blood vessel centerline image based on each blood vessel bifurcation point to obtain a blood vessel centerline segmented image.

Optionally, after the segmentation process is performed, the segment marks that are the same as those of each blood vessel segment in the first blood vessel segmented image can be used to mark each segmented blood vessel centerline segment, so far the obtained There is a one-to-one correspondence between the segmented image of the blood vessel centerline and the first segmented image of the blood vessel.

In this embodiment, after determining the blood vessel centerline image, the computer device may further determine at least one blood vessel bifurcation point based on the blood vessel centerline image and the first blood vessel segmented image; Each blood vessel centerline in the blood vessel centerline image is segmented to obtain a segmented image of the blood vessel centerline, that is, a segmented image of the blood vessel centerline corresponding to the first segmented image of the blood vessel is obtained, so that the user can be more intuitive. Master the blood vessel distribution of different blood vessels in different tissues or organs to improve user experience.

In an optional embodiment of the present application, before determining at least one blood vessel bifurcation point based on the blood vessel centerline image and the first blood vessel segment image, the computer device may further extend the blood vessel centerline image based on the initial black blood image For each blood vessel centerline in , the extended blood vessel centerline image is obtained; because the first blood vessel segmented image obtained after the blood vessel segmentation is performed on the initial black blood image using the first segmentation model, there may be a lack of edge blood vessels. , or using the above-mentioned longest two blood vessel segments to connect the center lines to obtain the blood vessel center line will lead to the incompleteness of the blood vessel center line. Therefore, in this embodiment, the two ends of each blood vessel in the blood vessel center line image are extended. After processing, the centerline of each blood vessel will be more complete after extended processing, and can better reflect the overall shape of the actual blood vessel in the initial black blood image.

Optionally, when extending each blood vessel centerline in the blood vessel centerline image, the two ends of each blood vessel centerline may be extended by a preset length based on the initial black blood image, or may be extended based on each blood vessel centerline. The blood vessel flow direction at both ends of the blood vessel center line is extended along the blood vessel flow direction by a preset length, etc. This embodiment does not limit the implementation manner and extension length of the blood vessel center line.

FIG. 9 is a schematic flowchart of a method for extracting a blood vessel centerline in another embodiment. This embodiment relates to an optional implementation process in which the computer device restores the blood vessel on the basis of the above-mentioned blood vessel centerline image, and further determines the target blood vessel image; as shown in FIG. 9 , on the basis of the above-mentioned embodiment Above, the above method also includes:

Step 901 , based on the initial black blood image, obtain a cross-sectional image corresponding to each point on the centerline of each blood vessel in the blood vessel centerline image.

Optionally, for each blood vessel centerline in the blood vessel centerline image, multiple points on the blood vessel centerline may be predetermined, and the distance between each point should be less than or equal to a preset distance threshold, for example: each point The distance between them can be less than or equal to 0.7mm; optionally, an interpolation algorithm can be used to perform dense processing on the centerline of each blood vessel, and the distance between the two corresponding to each blood vessel centerline is less than or equal to the preset distance threshold. points; then, based on the initial black blood image, a cross-sectional image of the blood vessel location corresponding to each point can be obtained.

Step 902: Process each cross-sectional image to obtain a target blood vessel image corresponding to the blood vessel centerline image.

Optionally, after obtaining the cross-sectional image corresponding to each point of the blood vessel centerline, the cross-sectional image corresponding to each point may be processed, so that the cross-sectional image corresponding to each point satisfies the preset blood vessel structure rules. , wherein the preset vascular structure rules may include, but are not limited to, the contour of the lumen does not exceed the contour of the vessel wall, the size of the lumen, the size of the vessel wall, the separation distance between the lumen and the vessel wall, etc.; When it is judged that the cross-sectional image corresponding to a certain point does not meet the rules of the blood vessel structure, the cross-sectional image corresponding to the point can be adjusted to obtain the adjusted cross-sectional image that meets the rules of the blood vessel structure; The target blood vessel image corresponding to the blood vessel centerline image can be generated from the cross-sectional images that satisfy the rules of blood vessel structure; the cross-sectional images of two adjacent points can be filled by filling to obtain the target blood vessel image.

In this embodiment, the computer device may acquire, based on the initial black blood image, cross-sectional images corresponding to each point on the centerline of each blood vessel in the blood vessel centerline image, and process each cross-sectional image to obtain a cross-sectional image corresponding to the blood vessel centerline The target blood vessel image corresponding to the image; the target blood vessel image can reflect the shape and distribution of each blood vessel in the initial black blood image, and can more intuitively display the blood vessel distribution of the object to be measured for the user, improving the user's viewing experience and visualization. .

FIG. 10 is a schematic flowchart of a method for extracting a blood vessel centerline in another embodiment. This embodiment involves an optional implementation process in which a computer device processes each cross-sectional image to obtain a target blood vessel image corresponding to the blood vessel centerline image; as shown in FIG. 10 , on the basis of the above embodiment , the above step 902 includes:

Step 1001: Input each cross-sectional image into the second segmentation model to obtain a segmentation result image corresponding to each cross-sectional image.

The segmentation result image corresponding to the cross-sectional image may be an annotated image with annotation information of the lumen contour and the vessel wall contour of the blood vessel in the cross-sectional image.

The second segmentation model is a segmentation model obtained by training a second initial segmentation network according to the cross-sectional sample images of multiple blood vessels and the segmentation result labels corresponding to the cross-sectional sample images of each blood vessel, and the second initial segmentation network can be It is based on an existing deep learning network of any type, and may also be a segmentation network combining multiple different types of networks, which is not limited in this embodiment. In addition, the second initial segmentation network may be the same as or different from the first initial segmentation network.

Step 1002: Based on the segmentation result image corresponding to each cross-sectional image, determine whether the blood vessel structure in each cross-sectional image satisfies a preset blood vessel structure rule.

Step 1003, in the case that the blood vessel structure in the cross-sectional image does not meet the preset blood vessel structure rules, remove the cross-sectional images that do not meet the blood vessel structure rules, and obtain a target blood vessel image according to the remaining cross-sectional images that meet the blood vessel structure rules .

Optionally, filling processing can be performed on the remaining cross-sectional images that satisfy the blood vessel structure rules to obtain the target blood vessel image; or based on the remaining cross-sectional images that satisfy the blood vessel structure rules, the first two adjacent cross-sectional images can be processed. Filling processing to generate an intermediate blood vessel image; in the intermediate blood vessel image, there is a blood vessel gap at the position where the cross-sectional image that does not meet the rules of the blood vessel structure is removed, and for the blood vessel gap, a second filling process can be further performed on the intermediate blood vessel image , to obtain the target blood vessel image; wherein, the first filling processing and the second filling processing may adopt different filling processing methods in the prior art.

Optionally, before the second filling process is performed on the intermediate blood vessel image to obtain the target blood vessel image, the missing blood vessel cross-sections in the intermediate blood vessel image (that is, the cross-sectional images that do not satisfy the blood vessel structure rules are removed), An interpolation algorithm can be used to determine a new blood vessel cross-section, and then the new blood vessel cross-section can be interpolated to the missing position in the intermediate blood vessel image to obtain an interpolated intermediate blood vessel image. Filling processing is performed to obtain the target blood vessel image.

Optionally, after the second filling process is performed on the interpolated intermediate blood vessel image, a post-processing operation may also be performed on the filled blood vessel image to obtain the target blood vessel image; optionally, the post-processing operation may include but It is not limited to removing non-vascular points on the blood vessel surface, smoothing the blood vessel surface, etc., to improve the smoothness of the blood vessel surface.

In this embodiment, the computer device obtains a segmentation result image corresponding to each cross-sectional image by inputting each cross-sectional image into the second segmentation model, and determines whether each cross-sectional image corresponds to the segmentation result image based on the segmentation result image corresponding to each cross-sectional image. Whether the blood vessel structure meets the preset blood vessel structure rules, if the blood vessel structure in the cross-sectional image does not meet the preset blood vessel structure rules, remove the cross-sectional image that does not meet the blood vessel structure rules, and according to the blood vessel structure rules The target blood vessel image is obtained by removing the remaining cross-sectional images that do not meet the rules of the blood vessel structure to generate the target blood vessel image by removing the cross-sectional images that do not meet the rules of the blood vessel structure, and the blood vessels in the obtained target blood vessel image are more consistent with For standard blood vessels, the obtained target blood vessel image is more accurate, which improves the accuracy of the target blood vessel image.

In an optional embodiment of the present application, based on the obtained multiple blood vessel bifurcation points and the above-mentioned target blood vessel image, further, the computer device may further, based on the respective blood vessel bifurcation points, analyze the target blood vessel image for the Each blood vessel in the target blood vessel is segmented to obtain a target blood vessel segmented image corresponding to the target blood vessel image; The standard blood vessel structure is more matched, the blood vessel surface is smoother, and the connectivity between each blood vessel segment with an associated relationship is better, and the blood vessel extraction effect is better.

FIG. 11 is a schematic flowchart of a method for extracting a blood vessel centerline in another embodiment. This embodiment relates to an optional implementation process for a computer device to perform tissue detection on an initial black blood image based on a blood vessel centerline image; as shown in FIG. 11 , on the basis of the foregoing embodiment, the foregoing method further includes: :

Step 1101 , based on the blood vessel centerline image, detect whether the target tissue exists in the blood vessel in the initial black blood image.

Optionally, in the first embodiment of the present application, after determining the blood vessel centerline image corresponding to the initial black blood image according to the initial black blood image, the computer device may further use the blood vessel centerline image to determine the object to be tested. The target tissue in the blood vessel is detected and located, for example, plaque detection or stenosis detection can be performed on the blood vessel of the object to be measured based on the blood vessel centerline image.

Optionally, based on the blood vessel centerline image, the blood vessels in the initial black blood image can be segmented, and the blood vessels in the segmented initial black blood image can be detected and analyzed to determine whether there is target tissue in the blood vessels; optional Specifically, after segmenting the blood vessels in the initial black blood image, the blood vessels can be analyzed layer by layer, and the cross-sectional images of the blood vessels corresponding to each layer can be obtained, and then, the cross-sectional images of the blood vessels in each layer can be obtained. The segmentation process is performed to determine the lumen and wall of the blood vessel, and then, the lumen and wall of the segmented blood vessel can be analyzed for the diameter of the blood vessel to identify the stenosis of the blood vessel, and the identified stenosis of the blood vessel. The components between the lumen and the wall of the tube are identified, and according to the identification results, it is judged whether there is a target tissue in the stenosis. It should be noted that, in this embodiment, the identification of the target tissue in the blood vessel may adopt the prior art, therefore, the specific implementation process of the identification of the target tissue will not be repeated here.

Step 1102 , in the case that the target tissue exists in the blood vessel in the initial black blood image, based on the initial black blood image, perform segmentation processing on the target tissue to obtain a segmentation result of the target tissue.

Optionally, when it is determined that the target tissue exists in the blood vessel in the initial black blood image, a preset third segmentation model can be used, and based on the initial black blood image, the target tissue is segmented to obtain the segmentation of the target tissue. As a result, the third segmentation model can be a segmentation model obtained by training the third initial segmentation network by using multiple black blood sample images with target tissues and target tissue labels corresponding to each black blood sample image. The initial segmentation network may be an existing deep learning network of any type, or may be a segmentation network combining multiple different types of networks, etc., which is not limited in this embodiment; in addition, the third initial segmentation network It may be the same as or different from the above-mentioned first initial segmentation network or the above-mentioned second initial segmentation network.

In this embodiment, the computer device detects whether the target tissue exists in the blood vessel in the initial black blood image based on the blood vessel centerline image; and in the case that the target tissue exists in the blood vessel in the initial black blood image, based on the initial black blood image, The target tissue is segmented to obtain the segmentation result of the target tissue; that is, the target tissue detection and segmentation process in this embodiment is based on the blood vessel centerline image obtained only by the detection and identification of the black blood image, and the initial black blood image is detected and analyzed. The detection and segmentation of the target tissue is realized. In other words, in the whole detection process, only the initial black blood image of the object to be tested can be obtained to achieve the extraction of the blood vessel centerline and the detection of the target tissue. The whole process is efficient and real-time. High precision.

The following will provide a complete embodiment of the extraction method of the blood vessel centerline, which may include the following steps:

1. Obtain an initial black blood image of the object to be tested, and input the initial black blood image into a preset first segmentation model to obtain a first blood vessel segment image corresponding to the initial black blood image and including at least one blood vessel segment , as shown in Figure 12(a);

2. For each blood vessel segment in the first blood vessel segment image, extract the center line of each blood vessel segment, as shown in Figure 12(b);

3. Based on the initial black blood image and the preset blood vessel connection strategy, connect the centerline of each blood vessel segment to obtain the blood vessel centerline image corresponding to the initial black blood image, as shown in Figure 12(c);

4. Extend the centerline of each blood vessel segment to obtain the extended blood vessel centerline image, as shown in Figure 12(d);

5. Determine at least one blood vessel bifurcation point based on the blood vessel centerline image and the first blood vessel segmentation image, and perform segmentation processing on each blood vessel centerline in the blood vessel centerline image based on each blood vessel bifurcation point to obtain a blood vessel. The segmented image of the centerline, as shown in Fig. 12(e);

6. Carry out the densification process on each blood vessel in the blood vessel centerline image to obtain multiple points on the centerline of each blood vessel, as shown in Figure 12(f), and then, based on the initial black blood image, obtain the blood vessel centerline Cross-sectional images corresponding to each point on the centerline of each blood vessel in the image, and determine whether the blood vessel structure in the cross-sectional image corresponding to each point satisfies the preset blood vessel structure rules; for cross-sectional images that do not meet the blood vessel structure rules Remove and retain the cross-sectional image that satisfies the rules of the blood vessel structure to obtain an intermediate blood vessel image, as shown in Figure 12(g);

7. For the missing blood vessel cross section in the intermediate blood vessel image, use an interpolation algorithm to determine the new blood vessel cross section corresponding to the missing position point, and interpolate the new blood vessel cross section to the missing position in the intermediate blood vessel image to obtain the interpolated middle blood vessel. The blood vessel image, and then, perform filling processing on the interpolated intermediate blood vessel image to obtain the target blood vessel image, as shown in Figure 12(h); The processed target blood vessel image is obtained, as shown in Figure 12(i);

8. Perform segmentation processing on the target blood vessel image based on each of the foregoing blood vessel bifurcation points to obtain a segmented image of the target blood vessel after branching, as shown in FIG. 12(j).

It should be noted that each schematic diagram in FIG. 12 above is only used for schematic illustration as an example of this embodiment, and is not used to limit the specific expression form of each step in this embodiment.

It should be understood that, although the steps in the flowcharts involved in the above embodiments are sequentially displayed according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, the execution of these steps is not strictly limited to the order, and these steps may be performed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily executed and completed at the same time, but may be performed at different times The execution order of these steps or phases is not necessarily sequential, but may be performed alternately or alternately with other steps or at least a part of the steps or phases in the other steps.

Based on the same inventive concept, an embodiment of the present application also provides a blood vessel centerline extraction device for implementing the above-mentioned blood vessel centerline extraction method. The solution to the problem provided by this device is similar to the solution described in the above method, so the specific definitions in the embodiments of one or more blood vessel centerline extraction devices provided below can refer to the above for the blood vessel centerline The limitation of the extraction method is not repeated here.

In one embodiment, as shown in FIG. 13, a device for extracting blood vessel centerline is provided, including: a first acquisition module 1301, a second acquisition module 1302 and a third acquisition module 1303, wherein:

The first acquisition module 1301 is used to acquire the initial black blood image of the object to be tested;

The second acquisition module 1302 is configured to input the initial black blood image into a preset first segmentation model to obtain a first blood vessel segmented image; wherein the first blood vessel segmented image includes at least one blood vessel segment;

The third acquisition module 1303 is configured to extract the centerline of each blood vessel segment in the first blood vessel segmented image to obtain a blood vessel centerline image corresponding to the initial black blood image.

In one embodiment, the third acquisition module 1303 includes an extraction unit and an acquisition unit; wherein, the extraction unit is used for extracting the centerline of each blood vessel segment for each blood vessel segment; the acquisition unit is used for The black blood image and the preset blood vessel connection strategy connect the centerline of each blood vessel segment to obtain the blood vessel centerline image corresponding to the initial black blood image.

In one of the embodiments, the above-mentioned extraction unit is specifically configured to determine the first end point and the second end point of the blood vessel segment for the blood vessel segment; based on the first blood vessel segment image, connect the first end point of the blood vessel segment endpoint and the second endpoint to obtain the centerline of the vessel segment.

In one of the embodiments, the above-mentioned extraction unit is specifically configured to determine the first endpoint and the second endpoint of each sub-vessel segment when the vessel segment includes multiple sub-vessel segments; and for each sub-vessel segment segment, based on the first blood vessel segment image, connect the first end point and the second end point of the sub-vessel segment to obtain the center line of the sub-vessel segment; based on the initial black blood image, connect the center lines of each sub-vessel segment, Obtain the centerline of the vessel segment.

In one embodiment, the device further includes a determination module and a fourth acquisition module; the determination module is configured to determine at least one blood vessel bifurcation point based on the blood vessel centerline image and the first blood vessel segmented image; the fourth acquisition module , which is used to segment each blood vessel centerline in the blood vessel centerline image based on each blood vessel bifurcation point to obtain a segmented image of the blood vessel centerline.

In one of the embodiments, the above determination module includes a first determination unit, a determination unit and a second determination unit; wherein the first determination unit is configured to determine the position of the first bifurcation point in the first blood vessel segmented image, and the position of the second bifurcation point in the blood vessel centerline image; the judgment unit is used for judging whether the position of the second bifurcation point is within the preset range of the position of the first bifurcation point; the second determination unit is used for When the position of the second bifurcation point is within a preset range of the position of the first bifurcation point, the second bifurcation point is determined as a blood vessel bifurcation point.

In one of the embodiments, the second determining unit is further configured to return to re-executing based on the initial black blood image and blood vessels when the position of the second bifurcation point is not within the preset range of the position of the first bifurcation point The connection strategy is to connect the centerlines of each blood vessel segment to obtain an adjusted blood vessel centerline image, and perform the steps of determining the position of the second bifurcation point in the adjusted blood vessel centerline image until the second bifurcation point until the position of the first bifurcation point is within the preset range of the position of the first bifurcation point.

In one of the embodiments, the apparatus further includes a fifth acquisition module and a sixth acquisition module; the fifth acquisition module is configured to acquire, based on the initial black blood image, each blood vessel centerline in the blood vessel centerline image. The cross-sectional image corresponding to the point; the sixth acquisition module is used for processing each cross-sectional image to obtain a target blood vessel image corresponding to the blood vessel centerline image.

In one of the embodiments, the above-mentioned sixth obtaining module includes a first obtaining unit, a judging unit and a second obtaining unit; wherein, the first obtaining unit is configured to input each cross-sectional image into the second segmentation model, respectively, to obtain each a segmentation result image corresponding to the cross-sectional image; a judging unit for judging whether the blood vessel structure in each cross-sectional image satisfies a preset blood vessel structure rule based on the segmentation result image corresponding to each cross-sectional image; a second acquiring unit for When the blood vessel structure in the cross-sectional image does not meet the preset blood vessel structure rule, the cross-sectional image that does not satisfy the blood vessel structure rule is removed, and the target blood vessel image is obtained according to the remaining cross-sectional images that satisfy the blood vessel structure rule.

In one of the embodiments, the device further includes a seventh acquisition module; the seventh acquisition module is configured to perform segmentation processing on each blood vessel in the target blood vessel image based on each blood vessel bifurcation point to obtain the target blood vessel image Corresponding target vessel segmented images.

In one of the embodiments, the device further includes a detection module and an eighth acquisition module; wherein, the detection module is used to detect whether there is a target tissue in the blood vessel in the initial black blood image based on the blood vessel centerline image; the eighth acquisition module , is used to perform segmentation processing on the target tissue based on the initial black blood image when the target tissue exists in the blood vessel in the initial black blood image to obtain the segmentation result of the target tissue.

Each module in the above-mentioned apparatus for extracting blood vessel centerline may be implemented in whole or in part by software, hardware and combinations thereof. The above modules can be embedded in or independent of the processor in the computer device in the form of hardware, or stored in the memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

In one embodiment, a computer device is provided, and the computer device may be a terminal, and its internal structure diagram may be as shown in FIG. 14 . The computer equipment includes a processor, memory, a communication interface, a display screen, and an input device connected by a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium, an internal memory. The nonvolatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the execution of the operating system and computer programs in the non-volatile storage medium. The communication interface of the computer equipment is used for wired or wireless communication with an external terminal, and the wireless communication can be realized by WIFI, mobile cellular network, NFC (Near Field Communication) or other technologies. The computer program implements a method for extracting a blood vessel centerline when executed by a processor. The display screen of the computer equipment may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment may be a touch layer covered on the display screen, or a button, a trackball or a touchpad set on the shell of the computer equipment , or an external keyboard, trackpad, or mouse.

Those skilled in the art can understand that the structure shown in FIG. 14 is only a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer equipment to which the solution of the present application is applied. Include more or fewer components than shown in the figures, or combine certain components, or have a different arrangement of components.

In one embodiment, a computer device is provided, including a memory and a processor, where a computer program is stored in the memory, and when the processor executes the computer program, the steps of the method for extracting a blood vessel centerline in any of the foregoing embodiments are implemented.

In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, implements the steps of the method for extracting a blood vessel centerline in any of the foregoing embodiments.

In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps of the method for extracting a blood vessel centerline in any of the foregoing embodiments.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage In the medium, when the computer program is executed, it may include the processes of the above-mentioned method embodiments. Wherein, any reference to a memory, a database or other media used in the various embodiments provided in this application may include at least one of a non-volatile memory and a volatile memory. Non-volatile memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive memory (ReRAM), magnetic variable memory (Magnetoresistive Random Memory) Access Memory, MRAM), Ferroelectric Random Access Memory (FRAM), Phase Change Memory (Phase Change Memory, PCM), graphene memory, etc. Volatile memory may include random access memory (Random Access Memory, RAM) or external cache memory, and the like. As an illustration and not a limitation, the RAM can be in various forms, such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM). The database involved in the various embodiments provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a blockchain-based distributed database, etc., but is not limited thereto. The processors involved in the various embodiments provided in this application may be general-purpose processors, central processing units, graphics processors, digital signal processors, programmable logic devices, data processing logic devices based on quantum computing, etc., and are not limited to this.

The technical features of the above embodiments can be combined arbitrarily. In order to make the description simple, all possible combinations of the technical features in the above embodiments are not described. However, as long as there is no contradiction in the combination of these technical features It is considered to be the range described in this specification.

The above-mentioned embodiments only represent several embodiments of the present application, and the descriptions thereof are relatively specific and detailed, but should not be construed as a limitation on the scope of the patent of the present application. It should be pointed out that for those skilled in the art, without departing from the concept of the present application, several modifications and improvements can be made, which all belong to the protection scope of the present application. Therefore, the scope of protection of the present application should be determined by the appended claims.

Claims (12)

1. a kind of extraction method of blood vessel centerline, is characterized in that, described method comprises: Obtain the initial black blood image of the object to be tested; Inputting the initial black blood image into a preset first segmentation model to obtain a first blood vessel segmented image; wherein the first blood vessel segmented image includes at least one blood vessel segment; The center line of each blood vessel segment in the first blood vessel segment image is extracted to obtain a blood vessel center line image corresponding to the initial black blood image. 2 . The method according to claim 1 , wherein the centerline of each blood vessel segment in the first blood vessel segment image is extracted to obtain a blood vessel centerline image corresponding to the initial black blood image. 3 . ,include: For each vessel segment, extract the centerline of each vessel segment; Based on the initial black blood image and a preset blood vessel connection strategy, the centerline of each blood vessel segment is connected to obtain a blood vessel centerline image corresponding to the initial black blood image. 3. The method according to claim 2, wherein the extracting the centerline of each blood vessel segment comprises: For the vessel segment, determining a first endpoint and a second endpoint of the vessel segment; Based on the first blood vessel segment image, the first end point and the second end point are connected to obtain the center line of the blood vessel segment. 4. The method according to claim 3, wherein the determining the first endpoint and the second endpoint of the blood vessel segment comprises: The blood vessel segment includes a plurality of sub-vessel segments; determining a first endpoint and a second endpoint of each of the sub-vessel segments; Correspondingly, connecting the first end point and the second end point based on the first blood vessel segment image to obtain the center line of the blood vessel segment, including: For each sub-vessel segment, based on the first blood vessel segment image, connect the first end point and the second end point of the sub-vessel segment to obtain the centerline of the sub-vessel segment; Based on the initial black blood image, the centerlines of the sub-vessel segments are connected to obtain the centerlines of the blood vessel segments. 5. The method according to claim 2, wherein the method further comprises: determining at least one vessel bifurcation point based on the vessel centerline image and the first vessel segment image; Based on each of the blood vessel bifurcation points, each blood vessel centerline in the blood vessel centerline image is segmented to obtain a segmented image of the blood vessel centerline. 6. The method according to claim 5, wherein the determining at least one blood vessel bifurcation point based on the blood vessel centerline image and the first blood vessel segment image comprises: determining the position of the first bifurcation point in the first blood vessel segmented image, and the position of the second bifurcation point in the blood vessel centerline image; judging whether the position of the second bifurcation point is within the preset range of the position of the first bifurcation point; If so, the second bifurcation point is determined as the blood vessel bifurcation point. 7 . The method according to claim 6 , wherein the method further comprises: if not, returning to re-execute the connection of the blood vessel segments based on the initial black blood image and the blood vessel connection strategy. 8 . obtain the adjusted centerline image of the blood vessel, and perform the step of determining the position of the second bifurcation point in the adjusted blood vessel centerline image, until the position of the second bifurcation point is located in the within the preset range of the position of the first bifurcation point. 8. The method according to claim 5, wherein the method further comprises: obtaining, based on the initial black blood image, a cross-sectional image corresponding to each point on the blood vessel centerline in the blood vessel centerline image; Each of the cross-sectional images is processed to obtain a target blood vessel image corresponding to the blood vessel centerline image. 9 . The method according to claim 8 , wherein the processing of each of the cross-sectional images to obtain a target blood vessel image corresponding to the blood vessel centerline image comprises: 10 . Inputting each of the cross-sectional images into the second segmentation model, respectively, to obtain a segmentation result image corresponding to each of the cross-sectional images; Based on the segmentation result image corresponding to each of the cross-sectional images, determine whether the blood vessel structure in each of the cross-sectional images satisfies a preset blood vessel structure rule; If not, remove the cross-sectional images that do not satisfy the blood vessel structure rule, and obtain the target blood vessel image according to the remaining cross-sectional images that satisfy the blood vessel structure rule. 10. The method according to claim 8, wherein the method further comprises: Based on each of the blood vessel bifurcation points, each blood vessel in the target blood vessel image is segmented to obtain a target blood vessel segmented image corresponding to the target blood vessel image. 11. The method of claim 1, wherein the method further comprises: based on the blood vessel centerline image, detecting whether there is a target tissue in the blood vessel in the initial black blood image; If yes, perform segmentation processing on the target tissue based on the initial black blood image to obtain a segmentation result of the target tissue. 12. A computer-readable storage medium on which a computer program is stored, characterized in that, when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 11 are implemented.

Images