WO2022095895A1

WO2022095895A1 - Vascular stenosis analysis method and apparatus

Info

Publication number: WO2022095895A1
Application number: PCT/CN2021/128410
Authority: WO
Inventors: 王思雯; 张番栋; 俞益洲; 李一鸣; 乔昕
Original assignee: 杭州深睿博联科技有限公司; 北京深睿博联科技有限责任公司
Priority date: 2020-11-05
Filing date: 2021-11-03
Publication date: 2022-05-12
Also published as: CN112288731A

Abstract

Disclosed is a vascular stenosis analysis method. The method comprises: firstly, determining, from a target heart CT angiography image, a target image area that comprises blood vessels and plaques; performing segmentation by using the target image area, so as to obtain a segmented blood vessel image and a segmented plaque image; and according to the segmented blood vessel image and the segmented plaque image, determining a target vascular stenosis analysis result corresponding to the target heart CT angiography image. In this way, since the plaque size is very small compared with blood vessels and a background, and a large amount of redundant information is removed from a target image area relative to a target heart CT angiography image, a pre-set segmentation model is not subjected to the interference of a large amount of redundant information in the target heart CT angiography image, and plaques and blood vessels in the target image area are accurately segmented, that is, the segmentation accuracy of plaques and blood vessels is improved, and therefore, the accuracy of a target vascular stenosis analysis result corresponding to the target heart CT angiography image can be improved.

Description

A kind of vascular stenosis analysis method and device

The present invention claims the priority of the application proposed by the applicant, the application date is November 5, 2020, the application number is CN2020112227115, and the title is "a method and device for analyzing vascular stenosis". The entire contents of the above application are incorporated herein by reference in their entirety.

technical field

The present application relates to the field of medical data analysis, and in particular, to a method and device for analyzing vascular stenosis.

Background of the Invention

Coronary heart disease is one of the most common causes of death in the world. Cardiac CT angiography (CCTA) is widely used in various countries as a non-invasive and highly sensitive examination method. The use of CCTA to determine the degree of stenosis of the diseased coronary lumen has important clinical value in assessing the severity of the disease and guiding treatment.

Most of the existing methods for determining the stenosis degree of the diseased coronary lumen directly estimate the stenosis degree of the lumen on the vascular surface reconstruction (CPR) image or straightened image, but due to the existing technology's impact on the plaque and blood vessels that cause vascular stenosis. The segmentation is not precise enough, so the estimated stenosis is not accurate enough. Therefore, how to accurately segment plaques and blood vessels is crucial for the analysis of vascular stenosis.

SUMMARY OF THE INVENTION

The present application provides a blood vessel stenosis analysis method and device, so as to improve the segmentation accuracy of plaques and blood vessels, thereby improving the accuracy of the target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image.

In a first aspect, the present application provides a method for analyzing vascular stenosis, the method comprising:

Obtain the target cardiac CT angiography image;

performing stretching and straightening processing on the target cardiac CT angiography image to obtain a processed cardiac CT angiography image;

determining a target image area in the processed cardiac CT angiography image, wherein the target image area includes blood vessels and plaque;

Inputting the target image area into a preset segmentation model to obtain segmented blood vessel images and plaque images;

According to the segmented blood vessel image and plaque image, the target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image is determined.

Optionally, performing stretching and straightening processing on the target cardiac CT angiography image to obtain a processed cardiac CT angiography image, including:

performing a curved surface reconstruction process on the target cardiac CT angiography image to obtain a curved surface reconstructed image, and using the curved surface reconstructed image as a processed cardiac CT angiography image; and/or,

Perform straightening imaging processing on the target cardiac CT angiography image to obtain a straightened image, and use the straightened image as a processed cardiac CT angiography image.

Optionally, the determining of the target image area in the processed cardiac CT angiography image includes:

Inputting the processed cardiac CT angiography image into a preset plaque detection model to obtain several candidate regions and the plaque probability corresponding to each candidate region, wherein the plaque probability corresponding to the candidate region reflects that the candidate region is a plaque the probability of a block;

Determine the target candidate region according to the respective corresponding patch probabilities of the several candidate regions;

According to the target candidate area, a target image area is determined; wherein, the target image area includes the target candidate area, and the area of the target image area is larger than the target candidate area.

Optionally, the preset segmentation model includes several downsampling layers and several upsampling layers;

Among them, the several downsampling layers are connected in cascade, and each downsampling layer is connected with an upsampling layer; and the feature maps obtained from all the upsampling layers with the same spatial resolution and the features obtained from the downsampling layers In the figure, a skip connection can be performed between the feature maps obtained by every two sampling layers; the input feature map and the output feature map of the preset segmentation model have the same spatial resolution.

The target cardiac CT angiography image is stretched and straightened according to N preset angles to obtain N processed cardiac CT angiography images; wherein, N is a positive integer;

Correspondingly, the determining of the target image region in the processed cardiac CT angiography image includes:

determining the target image regions corresponding to the N processed cardiac CT angiography images respectively;

Correspondingly, inputting the target image region into a preset segmentation model to obtain segmented blood vessel images and plaque images, including:

Inputting each target image area into a preset segmentation model, respectively, to obtain a segmented blood vessel image and a plaque image corresponding to each target image area;

Correspondingly, determining the blood vessel stenosis analysis result according to the segmented blood vessel image and plaque image, including:

Determine the blood vessel stenosis analysis results corresponding to each segmented blood vessel image and plaque image respectively;

The target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image is determined according to the blood vessel stenosis analysis result corresponding to each segmented blood vessel image.

Optionally, the vascular stenosis analysis result includes the stenosis rate of the vessel stenosis area and the stenosis degree of the vessel stenosis area; the target vessel stenosis analysis result includes the stenosis rate of the vessel stenosis area and the stenosis degree of the vessel stenosis area.

Optionally, the determining of the blood vessel stenosis analysis results corresponding to each of the segmented blood vessel images and the plaque images, respectively, includes:

For each segmented blood vessel image and plaque image, according to the blood vessel image and the plaque image, determine the diameter of the beginning end of the blood vessel, the diameter of the end end of the blood vessel, and the diameter of the narrowest part of the blood vessel; and, according to the diameter of the beginning end of the blood vessel, The diameter of the terminal end of the blood vessel and the diameter of the most stenotic point of the blood vessel determine the stenosis rate of the blood vessel image and the blood vessel stenosis region corresponding to the plaque image.

In a second aspect, the present application provides a blood vessel stenosis analysis device, the device comprising:

an acquisition unit for acquiring a target cardiac CT angiography image;

a processing unit, configured to perform stretching and straightening processing on the target cardiac CT angiography image to obtain a processed cardiac CT angiography image;

a determining unit, configured to determine a target image area in the processed cardiac CT angiography image, wherein the target image area includes blood vessels and plaques;

a segmentation unit, configured to input the target image region into a preset segmentation model to obtain segmented blood vessel images and plaque images;

An analysis unit, configured to determine a target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image according to the segmented blood vessel image and the plaque image.

Optionally, the processing unit is used for:

Optionally, the determining unit is used for:

Optionally, the processing unit is used for:

Perform stretching and straightening processing on the target cardiac CT angiography image according to N preset angles to obtain N processed cardiac CT angiography images; wherein, N is a positive integer;

Correspondingly, the determining unit is used for:

Correspondingly, the dividing unit is used for:

Correspondingly, the analysis unit is used to:

Optionally, the analysis unit is specifically used for:

In a third aspect, the present application provides a readable medium, including execution instructions, when a processor of an electronic device executes the execution instructions, the electronic device executes the method according to any one of the first aspects.

In a fourth aspect, the present application provides an electronic device, including a processor and a memory storing execution instructions. When the processor executes the execution instructions stored in the memory, the processor executes the first aspect. any of the methods described above.

It can be seen from the above technical solutions that, after obtaining the target cardiac CT angiography image, the present application can first perform stretching and straightening processing on the target cardiac CT angiography image to obtain the processed cardiac CT angiography image; then, determine the target cardiac CT angiography image. The target image area in the processed cardiac CT angiography image, wherein the target image area includes blood vessels and plaques; then, the target image area can be input into a preset segmentation model to obtain a segmented blood vessel image and the plaque image; then, the target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image may be determined according to the segmented blood vessel image and the plaque image. It can be seen that in the present application, a target image area including blood vessels and plaques is first determined from the target cardiac CT angiography image, and then the target image area is used for segmentation to obtain the segmented blood vessel image and plaque image. The segmented blood vessel image and plaque image are used to determine the target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image. In this way, since the size of the plaque is very small compared to the blood vessel and the background, and the target image area is relatively small relative to the target heart In terms of CT angiography images, a large amount of redundant information is removed, so the preset segmentation model can avoid the interference of a large amount of redundant information in the target cardiac CT angiography images, and accurately segment plaques and plaques in the target image area. Blood vessels, that is, the segmentation accuracy of plaques and blood vessels is improved, so that the accuracy of the target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image can be improved (for example, the accuracy of the stenosis rate of the blood vessel stenosis region can be improved).

Further effects of the above-mentioned non-conventional preferred mode will be described below in conjunction with specific embodiments.

Brief Description of Drawings

In order to explain the embodiments of the present application or the existing technical solutions more clearly, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments or the existing technology. Obviously, the accompanying drawings in the following description are only the For some embodiments described in the application, for those of ordinary skill in the art, other drawings can also be obtained based on these drawings without any creative effort.

1 is a schematic flowchart of a method for analyzing vascular stenosis of the present application;

FIG. 2 is a schematic structural diagram of a blood vessel stenosis analysis device according to an embodiment of the present application;

FIG. 3 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

MODES OF IMPLEMENTING THE INVENTION

In order to make the technical purpose, technical solutions and beneficial effects of the present invention clearer, the specific embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. The described specific embodiments are only a part of the embodiments of the present invention, not all of them Based on the specific embodiments of the present invention, other embodiments obtained by those skilled in the art without creative work shall fall within the protection scope of the present invention.

In order to make the purpose, technical solutions and advantages of the present invention clearer and more comprehensible, the present invention will be further described below in conjunction with the accompanying drawings and specific embodiments. Obviously, the described embodiments are only some, but not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

The inventors found that on cardiac CT angiography (CCTA) images, plaques and blood vessels were segmented, and the segmentation results were mapped onto CPR images and straightened images at multiple set angles to obtain CPR images and straightened images. Segmentation of plaques and blood vessels, or segmentation of plaques and blood vessels on CPR images and straightened images at multiple set angles, due to the very small size of the plaques compared to the blood vessels and background, these full-image-based methods are used. There is a lot of redundant information in the segmentation model, which cannot accurately segment plaques and blood vessels. According to the segmentation results, the diameter of the lumen cannot be accurately measured. Therefore, the calculated stenosis rate of the stenosis region of the blood vessel is inaccurate, and the stenosis rate corresponding to the stenosis rate is inaccurate. The degree is also inaccurate.

Therefore, the present application provides a method for analyzing blood vessel stenosis. Specifically, after the target cardiac CT angiography image can be acquired, the target cardiac CT angiography image can be stretched and straightened to obtain the processed cardiac CT blood vessel. angiography image; then, determine a target image area in the processed cardiac CT angiography image, wherein the target image area includes blood vessels and plaques; then, the target image area can be input into a preset segmentation model , to obtain the segmented blood vessel image and plaque image; then, the target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image may be determined according to the segmented blood vessel image and plaque image. It can be seen that in the present application, a target image area including blood vessels and plaques is first determined from the target cardiac CT angiography image, and then the target image area is used for segmentation to obtain the segmented blood vessel image and plaque image. The segmented blood vessel image and plaque image are used to determine the target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image. In this way, since the size of the plaque is very small compared to the blood vessel and the background, and the target image area is relatively small relative to the target heart In terms of CT angiography images, a large amount of redundant information is removed, so the preset segmentation model can avoid the interference of a large amount of redundant information in the target cardiac CT angiography images, and accurately segment plaques and plaques in the target image area. Blood vessels, that is, the segmentation accuracy of plaques and blood vessels is improved, so that the accuracy of the target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image can be improved (for example, the accuracy of the stenosis rate of the blood vessel stenosis region can be improved).

Various non-limiting embodiments of the present application will be described in detail below with reference to the accompanying drawings.

Referring to FIG. 1 , a method for analyzing blood vessel stenosis in an embodiment of the present application is shown, wherein the method can be completely applied to terminal devices (such as mobile phones, notebooks, electronic communication watches and other mobile devices), or can be completely applied to servers , or part of the steps may be applied to the terminal device, and part of the steps may be applied to the server. In this embodiment, the method may include the following steps, for example:

S101: Acquire a target cardiac CT angiography image.

In this embodiment, a target cardiac CT angiography (CCTA) image that needs to be analyzed for vascular stenosis may be referred to as a target cardiac CT angiography image. It should be noted that, under normal circumstances, the target cardiac CT angiography image may include blood vessel images, plaque images and other background images. It should be noted that the target cardiac CT angiography image may be obtained by scanning the patient with a scanning device, or may be sent and received by other devices. In this embodiment, the acquisition manner of the target cardiac CT angiography image is not limited.

S102: Perform stretching and straightening processing on the target cardiac CT angiography image to obtain a processed cardiac CT angiography image.

In order to acquire the blood vessel image more clearly, in this embodiment, after the target cardiac CT angiography image is acquired, the target cardiac CT angiography image may be stretched and straightened first, so as to obtain the processed cardiac CT angiography image. An angiography image, wherein the blood vessels in the processed cardiac CT angiography image are in a stretched and straightened state.

In this embodiment, stretching and straightening processing may be performed on the target cardiac CT angiography image by means of curved surface reconstruction and/or by means of straightening imaging processing. Next, these two methods will be introduced:

1. Stretch and straighten the target cardiac CT angiography image by means of curved surface reconstruction.

In this implementation manner, a curved surface reconstruction process may be performed on the target cardiac CT angiography image to obtain a curved surface reconstructed image, and the curved surface reconstructed image may be used as a processed cardiac CT angiography image.

It should be noted that the surface reconstruction (cerved projection reformation, CPR) technology is the extension and development of the MPR technology, that is, on the basis of MPR, a curve is drawn along the organ of interest (such as the blood vessels in the target cardiac CT angiography image), and the The volumetric metadata of the curve is reorganized in order to obtain CPR images. CPR technology can stretch and straighten twisted, shortened and overlapping structures such as blood vessels and colon, and display them in the same plane. It can display the whole course of the blood vessels along the coronary arteries.

2. Perform stretching and straightening processing on the target cardiac CT angiography image by means of straightening imaging processing.

In this embodiment, straightening imaging processing is performed on the target cardiac CT angiography image to obtain a straightened image, and the straightened image is used as a processed cardiac CT angiography image.

It should be noted that the straightening imaging process may be to draw a curve along the normal direction of the organ of interest (such as a blood vessel in a target cardiac CT angiography image), and reorganize the volume metadata along the curve to obtain a straightened image. Straightening imaging can stretch and straighten twisted, shortened, and overlapping structures such as blood vessels, colon, etc., and display them in the same plane.

It is understandable that both surface reconstruction (CPR) images and straightened images can display blood vessels in 3D on a 2D plane, but the imaging methods of the two are slightly different. So far, the method of performing stretching and straightening processing on the target cardiac CT angiography image through the method of curved surface reconstruction and/or the method of straightening imaging processing has been introduced.

It should be noted that when the plaque grows eccentrically, a slice cannot correctly reflect the size of the plaque, and sometimes the plaque is missed. Therefore, it is necessary to perform multi-angle slices on the target cardiac CT angiography image, and the processed cardiac CT angiography images at each angle can have corresponding vascular stenosis analysis results, so that the processed cardiac CT vessels from these angles can be integrated. The blood vessel stenosis analysis result corresponding to the angiography image is used to obtain the stenosis degree of the blood vessel in a certain lumen; specifically, in an implementation manner of this embodiment, the target cardiac CT angiography image can be analyzed according to N preset angles. Perform stretching and straightening processing to obtain N processed cardiac CT angiography images; wherein, N is a positive integer.

Specifically, the N preset angles may be preset according to actual requirements. For example, for a cardiac CT angiography image of 0-360 degrees, a processed cardiac CT angiography image may be reconstructed every 360/N degrees. N pieces of processed cardiac CT angiography images can be obtained, where N is a positive integer. It should be noted that, in this embodiment, the N processed cardiac CT angiography images may all be obtained by performing stretching and straightening processing on the target cardiac CT angiography image by means of curved surface reconstruction, or may be all obtained by stretching. It can be obtained by stretching and straightening the target cardiac CT angiography image by means of straight imaging, and it can also be obtained by stretching and straightening the target cardiac CT angiography image by means of curved surface reconstruction and straightening imaging processing. .

S103: Determine a target image region in the processed cardiac CT angiography image.

In this embodiment, a processed cardiac CT angiography image is obtained, and a target image area in the processed cardiac CT angiography image can be determined, wherein the target image area includes blood vessels and plaques. It can be understood that the target image area is a part of the image area in the processed cardiac CT angiography image, and compared with the processed cardiac CT angiography image, the target image area has less redundant information in the target image area. Therefore, the segmentation of plaques and blood vessels based on the target image area can make the preset segmentation model avoid the interference of a large amount of redundant information in the target cardiac CT angiography image, and accurately segment the plaques and blood vessels in the target image area. That is, the segmentation accuracy of plaques and blood vessels is improved, so that the accuracy of the target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image can be improved (for example, the accuracy of the stenosis rate of the blood vessel stenosis area is improved).

It should be noted that, in this embodiment, the method of determining the target image region in the processed cardiac CT angiography image may be an image processing method or a neural network processing method. The following will take the neural network processing method as an example to introduce.

Specifically, in an implementation manner of this embodiment, the processed cardiac CT angiography image may be input into a preset plaque detection model to obtain several candidate regions and the plaque probability corresponding to each candidate region, Among them, the patch probability corresponding to the candidate area reflects the probability that the candidate area is a patch. It should be emphasized that if the patch probability corresponding to the candidate area is higher, the probability that the candidate area is an image corresponding to a patch is higher. On the contrary, if the probability of the patch corresponding to the candidate area is lower, it means that the probability that the candidate area is the image corresponding to the patch is lower. It should be noted that the preset plaque detection model may be a deep convolutional neural network trained based on several training image groups, wherein each training image group includes a training image, a labeled training image corresponding to the training image, and the The labeled training image is an image obtained by labeling the patches with candidate frames on the basis of the training image.

Then, the target candidate area may be determined according to the respective corresponding patch probabilities of the several candidate areas. It should be noted that the target candidate area may be zero or one or more. In an implementation manner, it may be Identify multiple target candidate regions. For example, it is assumed that a candidate area with a patch probability greater than a preset threshold in several candidate areas can be used as the target candidate area, or, several candidate areas can be sorted according to the patch probability from high to low, and the rank is higher than the predetermined threshold. Let the ranked candidate region be the target candidate region.

Next, a target image area may be determined according to the target candidate area. Wherein, the target image area includes the target candidate area, and the area of the target image area is larger than the target candidate area. For example, the target candidate area can be taken as the center, the preset fixed size can be expanded around, and the image blocks can be cut to obtain the target image area, wherein the target image area includes the plaque image, the blood vessel image and the background image; it needs to be explained Yes, considering that on processed cardiac CT angiography images (such as CPR images) with multiple set angles, the morphological changes of plaques are very large. In order to avoid the loss of plaque information caused by image scaling operations, a preset fixed The aspect ratio (one implementation is to preset a fixed size) cuts the image block to obtain the target image area, that is, it can avoid the loss of patch information caused by the image scaling operation.

It should be noted that, in an implementation manner of this embodiment, when N processed cardiac CT angiography images are acquired, the respective targets corresponding to the N processed cardiac CT angiography images may be determined. The image area, that is, the target image area corresponding to each processed cardiac CT angiography image can be determined respectively. The method for determining the respective corresponding target image regions of each processed cardiac CT angiography image may refer to the relevant description of S103, which will not be repeated here.

S104: Input the target image region into a preset segmentation model to obtain segmented blood vessel images and plaque images.

After the target image region in the processed cardiac CT angiography image is determined, the target image region can be input into a preset segmentation model to obtain segmented blood vessel images and plaque images. It should be noted that the preset segmentation model may be a deep convolutional neural network trained based on several segmentation training image groups, such as PSPNet, DeepLabV3, UNet, and so on. Wherein, each group of segmentation training images may include a training image to be segmented, and blood vessel images and plaque images corresponding to the training images to be segmented. It should be noted that compared with the processed cardiac CT angiography image, the target image area has less redundant information, and the preset segmentation model is based on the target image area for plaques and blood vessels. Segmentation is performed instead of the full image of the processed cardiac CT angiography image. Therefore, compared to the full image, the preset segmentation model based on the segmentation of image blocks (ie, target image regions) is semantically simpler and has a more fixed structure, and The preset segmentation model can avoid the interference of a large amount of redundant information in the target cardiac CT angiography image, so that the plaque image and blood vessel image in the target image area can be accurately segmented, and more accurate plaque and blood vessel segmentation results can be obtained. The segmentation accuracy of plaques and blood vessels is improved, thereby improving the accuracy of the target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image (for example, improving the accuracy of the stenosis rate of the blood vessel stenosis region).

It should be noted that, in an implementation manner, due to the simple semantics and fixed structure of the image block (that is, the target image region), its high-resolution shallow features and low-resolution deep features are equally important. Therefore, It is possible to use both high-resolution shallow features and low-resolution deep features for segmentation image learning. Therefore, in the preset segmentation model of this embodiment, not only can the feature map be up-sampled when the feature map with the smallest spatial resolution is down-sampled, but also the feature map can be up-sampled when the feature map is down-sampled to the middle. Sampling, and skip connections can be made between feature maps of the same spatial resolution, so that the features of different levels can be fully integrated, so that the preset segmentation model can learn more detailed information, and can learn more comprehensive information.

Specifically, in this implementation manner, the preset segmentation model may include several down-sampling layers and several up-sampling layers; wherein, the several down-sampling layers are cascade-connected (that is, the down-sampling layers). are connected one by one, such as including three downsampling layers, downsampling layer A is connected with downsampling layer B, and downsampling layer B is connected with downsampling layer C), and each downsampling layer is connected with an upsampling layer; and In the feature maps obtained by all the up-sampling layers and the feature maps obtained by the down-sampling layers with the same spatial resolution, skip connections can be made between the feature maps obtained by every two sampling layers; the input features of the preset segmentation model The spatial resolution of the map and the output feature map is the same, that is, the spatial resolution of the input feature map and the output feature map of the network is the same; The feature map obtained by the upsampling layer a and the feature map obtained by the upsampling layer b, the feature map obtained by the downsampling layer A and the feature map obtained by the upsampling layer a and the feature map obtained by the upsampling layer b can be compared. For skip connections, skip connections can also be made between the feature map obtained by the upsampling layer a and the feature map obtained by the upsampling layer b.

It should be noted that, in an implementation manner of this embodiment, when the respective target image regions corresponding to the N processed cardiac CT angiography images are obtained, each target image region may be input into the preset segmentation respectively. The model is used to obtain the segmented blood vessel images and plaque images corresponding to each target image area respectively. The manner of determining the segmented blood vessel image and the plaque image corresponding to each target image region respectively refer to the relevant description of S104, and details are not repeated here.

S105: Determine the target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image according to the segmented blood vessel image and the plaque image.

In this embodiment, after the segmented blood vessel image and plaque image corresponding to the target cardiac CT angiography image are determined, the target cardiac CT angiography image may be determined according to the segmented blood vessel image and plaque image Corresponding target vessel stenosis analysis results. Among them, the target vascular stenosis analysis results include the stenosis rate of the vascular stenosis area and the stenosis degree of the vascular stenosis area; the stenosis rate of the vascular stenosis area can be understood as the stenosis ratio of the blood vessel, the stenosis rate of the vascular stenosis area and the stenosis degree of the vascular stenosis area can be Reflects the severity of stenosis in the stenotic area of the blood vessel.

Wherein, in this embodiment, the method for determining the stenosis rate of the blood vessel stenosis area is: first, according to the blood vessel image and plaque image corresponding to a target image area, determine the diameter of the starting end of the blood vessel, the diameter of the ending end of the blood vessel, and the diameter of the blood vessel in the target image area. The diameter of the most stenotic place; then, according to the diameter of the beginning end of the blood vessel, the diameter of the end end of the blood vessel, and the diameter of the most stenotic place of the blood vessel, determine the stenosis rate of the blood vessel stenosis region corresponding to the blood vessel image and the plaque image, for example, First, the diameters of the two ends (the diameter of the beginning end of the blood vessel, the diameter of the end end of the blood vessel) are weighted and summed to obtain the reference diameter, and then the formula: (reference diameter - diameter of the most stenotic blood vessel)/reference diameter to calculate the stenosis area of the blood vessel. The stenosis rate, that is, the ratio of the difference between the reference diameter and the diameter at the most stenotic point of the blood vessel to the reference diameter, is taken as the stenosis rate of the stenosis area of the blood vessel.

It should be noted that, in an implementation manner of this embodiment, when the segmented blood vessel images and plaque images corresponding to each target image region are obtained, the segmented blood vessel images and plaque images may be determined first. Each image corresponds to a blood vessel stenosis analysis result, wherein the blood vessel stenosis analysis result may include the stenosis rate of the blood vessel stenosis region and the stenosis degree of the blood vessel stenosis region. For example, for each segmented blood vessel image and plaque image, according to the blood vessel image and the plaque image, determine the diameter of the beginning end of the blood vessel, the diameter of the end end of the blood vessel, and the diameter of the narrowest part of the blood vessel; and, according to the beginning end of the blood vessel The diameter of the blood vessel, the diameter of the end of the blood vessel, and the diameter of the most stenotic part of the blood vessel, to determine the stenosis rate of the blood vessel image and the blood vessel stenosis area corresponding to the plaque image. The reference diameter is obtained by the weighted sum of the diameter of the terminal end of the blood vessel), and then the stenosis rate of the stenotic area of the blood vessel is calculated by the formula: (reference diameter - the diameter of the most stenotic part of the blood vessel)/reference diameter, that is, the reference diameter and the diameter of the most stenotic part of the blood vessel are calculated. The ratio of the difference to the reference diameter was used as the stenosis rate of the stenotic area of the vessel. One measure of quantitative information embodied by the stenosis rate includes information on the effect on the determined local vascular blood flow.

Next, the target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image may be determined according to the blood vessel stenosis analysis results corresponding to each of the segmented blood vessel images. For example, the maximum value, the average value or the weighted average value of the stenosis ratios of the vascular stenosis regions corresponding to the respective segmented vascular images can be used as the stenosis ratios of the vascular stenosis regions in the target vascular stenosis analysis result. The corresponding relationship between the stenosis rate and the stenosis degree is used to determine the stenosis degree corresponding to the stenosis rate of the vascular stenosis area in the target vascular stenosis analysis result; or the corresponding relationship between the preset stenosis rate and the stenosis degree can also be used to determine each segmented vascular image. The respective corresponding stenosis degrees, and then the maximum value, average value or weighted average value of the stenosis degrees among several stenosis degrees is used as the stenosis degree of the vascular stenosis region in the target vascular stenosis analysis result.

It can be seen from the above technical solutions that, after obtaining the target cardiac CT angiography image, the present application can first perform stretching and straightening processing on the target cardiac CT angiography image to obtain the processed cardiac CT angiography image; then, determine the target cardiac CT angiography image. The target image area in the processed cardiac CT angiography image, wherein the target image area includes blood vessels and plaques; then, the target image area can be input into a preset segmentation model to obtain a segmented blood vessel image and the plaque image; then, the target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image may be determined according to the segmented blood vessel image and the plaque image. It can be seen that in the present application, a target image area including blood vessels and plaques is first determined from the target cardiac CT angiography image, and then the target image area is used for segmentation to obtain the segmented blood vessel image and plaque image. The segmented blood vessel image and plaque image are used to determine the target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image. In this way, since the size of the plaque is very small compared to the blood vessel and the background, and the target image area is relatively small relative to the target heart In terms of CT angiography images, a large amount of redundant information is removed, so the preset segmentation model can avoid the interference of a large amount of redundant information in the target cardiac CT angiography images, and accurately segment plaques and plaques in the target image area. Blood vessels, that is, the segmentation accuracy of plaques and blood vessels is improved, so that the accuracy of the target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image can be improved (for example, the accuracy of the stenosis rate of the blood vessel stenosis region can be improved). That is to say, the present application uses a preset segmentation model based on image blocks (ie, target image areas) to avoid the loss of plaque information, fully fuse features at different levels, and obtain accurate plaques and blood vessels in the vascular stenosis area. Segment the image, accurately measure the diameter of the vascular lumen according to the segmented image, calculate the stenosis rate of the vascular lumen, and obtain the degree of stenosis corresponding to the stenosis rate, which is conducive to more accurate assessment of the severity of the disease and better guidance treat. The solution of the present application is fully realized automatically, and operations such as manual positioning of the vascular stenosis area and interactive measurement of the diameter are not required, which greatly saves manpower.

As shown in FIG. 2 , it is a specific embodiment of a blood vessel stenosis analysis device described in this application. The apparatus described in this embodiment is a physical apparatus for executing the method described in the foregoing embodiment. The technical solutions thereof are essentially the same as those of the above-mentioned embodiments, and the corresponding descriptions in the above-mentioned embodiments are also applicable to this embodiment. The device described in this embodiment includes:

an acquisition unit 501, configured to acquire a target cardiac CT angiography image;

a processing unit 502, configured to perform stretching and straightening processing on the target cardiac CT angiography image to obtain a processed cardiac CT angiography image;

a determining unit 503, configured to determine a target image area in the processed cardiac CT angiography image, wherein the target image area includes blood vessels and plaques;

A segmentation unit 504, configured to input the target image region into a preset segmentation model to obtain segmented blood vessel images and plaque images;

The analyzing unit 505 is configured to determine the target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image according to the segmented blood vessel image and the plaque image.

Optionally, the processing unit 502 is configured to:

Optionally, the determining unit 503 is configured to:

Optionally, the processing unit 502 is configured to:

Correspondingly, the determining unit 503 is used for:

Correspondingly, the dividing unit 504 is used for:

Correspondingly, the analysis unit 505 is used for:

Optionally, the analysis unit 505 is specifically used for:

FIG. 3 is a schematic structural diagram of an electronic device provided by an embodiment of the present application. At the hardware level, the electronic device includes a processor, and optionally an internal bus, a network interface, and a memory. The memory may include memory, such as high-speed random-access memory (Random-Access Memory, RAM), or may also include non-volatile memory (non-volatile memory), such as at least one disk memory. Of course, the electronic equipment may also include hardware required for other services.

The processor, network interface and memory can be connected to each other through an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect, peripheral component interconnect standard) bus or an EISA (Extended Component Interconnect) bus. Industry Standard Architecture, extended industry standard structure) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, only one bidirectional arrow is used in FIG. 3, but it does not mean that there is only one bus or one type of bus.

Memory for storing execution instructions. Specifically, a computer program that executes instructions can be executed. The memory may include memory and non-volatile memory and provide instructions and data for execution to the processor.

In a possible implementation manner, the processor reads the corresponding execution instructions from the non-volatile memory into the memory and then executes the execution, and also obtains the corresponding execution instructions from other devices, so as to form vascular stenosis at the logical level Analytical device. The processor executes the execution instructions stored in the memory, so as to implement the blood vessel stenosis analysis method provided in any embodiment of the present application through the executed execution instructions.

The above-mentioned method performed by the blood vessel stenosis analysis apparatus provided in the embodiment shown in FIG. 1 of the present application may be applied to a processor, or implemented by a processor. A processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above-mentioned method can be completed by a hardware integrated logic circuit in a processor or an instruction in the form of software. The above-mentioned processor can be a general-purpose processor, including a central processing unit (Central Processing Unit, CPU), a network processor (Network Processor, NP), etc.; it can also be a digital signal processor (Digital Signal Processor, DSP), dedicated integrated Circuit (Application Specific Integrated Circuit, ASIC), Field Programmable Gate Array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The methods, steps, and logic block diagrams disclosed in the embodiments of this application can be implemented or executed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

The steps of the method disclosed in conjunction with the embodiments of the present application may be directly embodied as executed by a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules may be located in random access memory, flash memory, read-only memory, programmable read-only memory or electrically erasable programmable memory, registers and other storage media mature in the art. The storage medium is located in the memory, and the processor reads the information in the memory, and completes the steps of the above method in combination with its hardware.

An embodiment of the present application also provides a readable medium, where an execution instruction is stored in the readable storage medium, and when the stored execution instruction is executed by a processor of an electronic device, the electronic device can be enabled to execute the execution of the instructions provided in any embodiment of the present application. The vascular stenosis analysis method is specifically used to perform the above-mentioned vascular stenosis analysis method.

The electronic device described in each of the foregoing embodiments may be a computer.

As will be appreciated by those skilled in the art, the embodiments of the present application may be provided as a method or a computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or a combination of software and hardware.

Each embodiment in this application is described in a progressive manner, and the same and similar parts between the various embodiments may be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the apparatus embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and reference may be made to some descriptions of the method embodiments for related parts.

It should also be noted that the terms "comprising", "comprising" or any other variation thereof are intended to encompass a non-exclusive inclusion such that a process, method, article or device comprising a series of elements includes not only those elements, but also Other elements not expressly listed, or which are inherent to such a process, method, article of manufacture, or apparatus are also included. Without further limitation, an element qualified by the phrase "comprising a..." does not preclude the presence of additional identical elements in the process, method, article of manufacture, or device that includes the element.

The above descriptions are merely examples of the present application, and are not intended to limit the present application. Various modifications and variations of this application are possible for those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included within the scope of the claims of the present application.

Industrial Applicability

The blood vessel stenosis analysis method and device provided by the present invention utilize image recognition technology combined with segmentation model to determine the location of blood vessels and plaques and quantify stenosis information, and make full use of the automatic processing basis of computer technology, so that the massive cardiac CT angiography images can be analyzed in The efficiency of localization and assessment of stenotic vessels is greatly improved. The formed products can be mass-produced and quickly applied to systems or scenarios with high demand for the diagnosis of diseased coronary arteries.

Claims

A vascular stenosis analysis method, characterized in that the method comprises:

Obtain the target cardiac CT angiography image;

performing stretching and straightening processing on the target cardiac CT angiography image to obtain a processed cardiac CT angiography image;

determining a target image area in the processed cardiac CT angiography image, wherein the target image area includes blood vessels and plaques;

Inputting the target image area into a preset segmentation model to obtain segmented blood vessel images and plaque images;

According to the segmented blood vessel image and plaque image, the target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image is determined.
The method according to claim 1, characterized in that, performing stretching and straightening processing on the target cardiac CT angiography image to obtain a processed cardiac CT angiography image, comprising:

performing a curved surface reconstruction process on the target cardiac CT angiography image to obtain a curved surface reconstructed image, and using the curved surface reconstructed image as a processed cardiac CT angiography image; and/or,

Perform straightening imaging processing on the target cardiac CT angiography image to obtain a straightened image, and use the straightened image as a processed cardiac CT angiography image.
The method according to claim 1, wherein the determining the target image region in the processed cardiac CT angiography image comprises:

Inputting the processed cardiac CT angiography image into a preset plaque detection model to obtain several candidate regions and the plaque probability corresponding to each candidate region, wherein the plaque probability corresponding to the candidate region reflects that the candidate region is a plaque block probability;

Determine the target candidate region according to the respective corresponding patch probabilities of the several candidate regions;

According to the target candidate area, a target image area is determined; wherein, the target image area includes the target candidate area, and the area of the target image area is larger than the target candidate area.
The method according to claim 1, wherein the preset segmentation model comprises several down-sampling layers and several up-sampling layers;

Among them, the several downsampling layers are connected in cascade, and each downsampling layer is connected with an upsampling layer; and the feature maps obtained from all the upsampling layers with the same spatial resolution and the features obtained from the downsampling layers In the figure, a skip connection can be performed between the feature maps obtained by every two sampling layers; the input feature map and the output feature map of the preset segmentation model have the same spatial resolution.
The method according to any one of claims 1-4, wherein the performing stretching and straightening processing on the target cardiac CT angiography image to obtain a processed cardiac CT angiography image, comprising:

Perform stretching and straightening processing on the target cardiac CT angiography image according to N preset angles to obtain N processed cardiac CT angiography images; wherein, N is a positive integer;

Correspondingly, the determining of the target image region in the processed cardiac CT angiography image includes:

determining the target image regions corresponding to the N processed cardiac CT angiography images respectively;

Correspondingly, inputting the target image region into a preset segmentation model to obtain segmented blood vessel images and plaque images, including:

Inputting each target image area into a preset segmentation model, respectively, to obtain a segmented blood vessel image and a plaque image corresponding to each target image area;

Correspondingly, determining the blood vessel stenosis analysis result according to the segmented blood vessel image and plaque image, including:

Determine the blood vessel stenosis analysis results corresponding to each segmented blood vessel image and plaque image respectively;

The target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image is determined according to the blood vessel stenosis analysis result corresponding to each segmented blood vessel image.
The method according to claim 5, wherein the vascular stenosis analysis result includes the stenosis rate of the vascular stenosis area and the stenosis degree of the vascular stenosis area; the target vascular stenosis analysis result includes the stenosis rate of the vascular stenosis area and the stenosis degree of the vessel How narrow the narrow area is.
The method according to claim 5 or 6, wherein the determining of the blood vessel stenosis analysis results corresponding to each of the segmented blood vessel images and the plaque images, respectively, comprises:

For each segmented blood vessel image and plaque image, according to the blood vessel image and the plaque image, determine the diameter of the beginning end of the blood vessel, the diameter of the end end of the blood vessel, and the diameter of the narrowest part of the blood vessel; and, according to the diameter of the beginning end of the blood vessel, The diameter of the terminal end of the blood vessel and the diameter of the most stenotic point of the blood vessel determine the stenosis rate of the blood vessel image and the blood vessel stenosis region corresponding to the plaque image.
A readable medium, characterized in that the readable medium includes execution instructions, and when a processor of an electronic device executes the execution instructions, the electronic device executes the method according to any one of claims 1-7 .
An electronic device, characterized in that the electronic device includes a processor and a memory storing execution instructions, and when the processor executes the execution instructions stored in the memory, the processor executes the execution as claimed in claim 1 The method of any of -7.
A vascular stenosis analysis device, characterized in that the device comprises:

an acquisition unit for acquiring a target cardiac CT angiography image;

a processing unit, configured to perform stretching and straightening processing on the target cardiac CT angiography image to obtain a processed cardiac CT angiography image;

a determining unit, configured to determine a target image area in the processed cardiac CT angiography image, wherein the target image area includes blood vessels and plaques;

a segmentation unit, configured to input the target image region into a preset segmentation model to obtain segmented blood vessel images and plaque images;

An analysis unit, configured to determine a target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image according to the segmented blood vessel image and the plaque image.
The vascular stenosis analysis device according to claim 10, wherein the processing unit is further configured to perform a curved surface reconstruction process on the target cardiac CT angiography image to obtain a curved surface reconstructed image, and then convert the curved surface reconstructed image to the curved surface reconstruction image. as processed cardiac CT angiography images; and/or,

Perform straightening imaging processing on the target cardiac CT angiography image to obtain a straightened image, and use the straightened image as a processed cardiac CT angiography image.
The vascular stenosis analysis device according to claim 10, wherein the determining unit is further configured to input the processed cardiac CT angiography image into a preset plaque detection model to obtain several candidate regions and each The patch probability corresponding to the candidate area, wherein the patch probability corresponding to the candidate area reflects the probability that the candidate area is a patch;

Determine the target candidate region according to the respective corresponding patch probabilities of the several candidate regions;

According to the target candidate area, a target image area is determined; wherein, the target image area includes the target candidate area, and the area of the target image area is larger than the target candidate area.
The blood vessel stenosis analysis device according to claim 10, wherein the preset segmentation model comprises several down-sampling layers and several up-sampling layers;

Among them, the several downsampling layers are connected in cascade, and each downsampling layer is connected with an upsampling layer; and the feature maps obtained from all the upsampling layers with the same spatial resolution and the features obtained from the downsampling layers In the figure, a skip connection can be performed between the feature maps obtained by every two sampling layers; the input feature map and the output feature map of the preset segmentation model have the same spatial resolution.
The blood vessel stenosis analysis device according to any one of claims 10 to 13, wherein the processing unit is further configured to perform stretching and straightening processing on the target cardiac CT angiography image according to N preset angles, to obtain N processed cardiac CT angiography images; wherein, N is a positive integer;

The determining unit is further configured to determine the target image regions corresponding to the N processed cardiac CT angiography images respectively;

The segmentation unit is further configured to input each target image region into a preset segmentation model, respectively, to obtain segmented blood vessel images and plaque images corresponding to each target image region respectively;

The analysis unit is further configured to determine the blood vessel stenosis analysis results corresponding to each of the segmented blood vessel images and the plaque images respectively;

The target blood vessel stenosis analysis result corresponding to the target cardiac CT angiography image is determined according to the blood vessel stenosis analysis result corresponding to each segmented blood vessel image.
The vascular stenosis analysis device according to claim 14, wherein the vascular stenosis analysis result comprises the stenosis rate of the vascular stenosis region and the stenosis degree of the vascular stenosis region; the target vascular stenosis analysis result comprises the stenosis of the vascular stenosis region rate and the degree of stenosis in the stenotic area of the vessel.
The blood vessel stenosis analysis device according to claim 14 or 15, wherein the analysis unit is further configured to, for each segmented blood vessel image and plaque image, determine the blood vessel image and the plaque image according to the blood vessel image and the plaque image. the diameter of the starting end of the blood vessel, the diameter of the ending end of the blood vessel, and the diameter of the most stenotic place of the blood vessel; and, according to the diameter of the starting end of the blood vessel, the diameter of the ending end of the blood vessel, and the diameter of the most stenotic place of the blood vessel, determine the blood vessel image and the plaque image The stenosis rate of the corresponding vascular stenosis area.