WO2020019739A1

WO2020019739A1 - Vascular wall plaque segmentation method and apparatus, and computer-readable storage medium

Info

Publication number: WO2020019739A1
Application number: PCT/CN2019/078891
Authority: WO
Inventors: 郑海荣; 刘新; 胡战利; 张娜; 梁栋; 杨永峰
Original assignee: 深圳先进技术研究院
Priority date: 2018-07-24
Filing date: 2019-03-20
Publication date: 2020-01-30
Also published as: CN109285158A

Abstract

A vascular wall plaque segmentation method and apparatus, and a computer-readable storage medium. The vascular wall plaque segmentation method comprises: obtaining a vascular wall image; iteratively performing N times of down-sampling convolution processing and first feature information extraction for the vascular wall image; if the iterative process of extracting the first feature information for N times is completed, performing N times of up-sampling convolution processing and second feature information extraction for an output object of the N-th down-sampling convolution processing based on the first feature information extracted for the N-th time; if the iterative process of extracting the second feature information for N times is completed, performing plaque segmentation on the vascular wall image based on a pre-trained classifier to determine whether plaques exist in the vascular wall image based on the segmentation result. The method can effectively improve the recognition efficiency of the plaque of the vascular wall.

Description

Vessel wall plaque segmentation method and device, and computer-readable storage medium

Technical field

The present application relates to the field of biomedicine, and in particular, to a method, a device, and a computer-readable storage medium for segmenting blood vessel wall plaques.

Background technique

Atherosclerosis (atherosclerosis, AS) is the main cause of cardiovascular and cerebrovascular diseases such as coronary heart disease, cerebral infarction and peripheral vascular disease. The cause of the disease is that cholesterol, fat-like, sugars and other substances accumulate in the large and middle arteries (for example: carotid arteries) to form plaques and block blood vessels, so that the tissues or organs supplied by the artery will be ischemic or necrotic. In addition, plaques that block blood vessels may also fall off. Once the plaques fall off the vessel wall, it may cause stroke and even death. Since cardiovascular and cerebrovascular diseases have become the number one killer of human health, studying atherosclerosis and its diagnostic measures is an extremely important task in medical research.

At present, magnetic resonance imaging (MRI) technology is used to obtain scanned images of arterial cross sections (that is, blood vessel wall images), and then a series of pictures of the same patient are carefully viewed and applied by specially trained medical staff or medical experts. Expertise identifies whether there are plaques on the vessel wall in the image of the vessel wall, so that the patient's risk level of cardiovascular and cerebrovascular disease can be initially diagnosed.

Because the blood vessel wall images taken for the same patient often contain multiple images, the method of manual judgment by medical staff or medical experts is not only time-consuming, but also has poor reproducibility, and is easily affected by doctors' experience and subjective factors.

Summary of the Invention

The present application provides a method, a device, and a computer-readable storage medium for segmenting a blood vessel wall plaque, which can be used to improve the recognition efficiency of a blood vessel wall plaque.

A first aspect of the present application provides a plaque segmentation method for a blood vessel wall, including:

Obtain a blood vessel wall image;

Performing down-sampling convolution processing on the blood vessel wall image;

Extracting first feature information, where the first feature information is feature information of an output object of a recent down-sampling convolution process;

If the iterative process of extracting the first feature information is not completed N times, the output object of the latest down-sampling convolution process is subjected to down-sampling convolution processing based on the currently extracted first feature information, and then the extracting the first feature is performed iteratively Information steps

If the iterative process of extracting the first feature information is performed N times, based on the first feature information extracted at the Nth time, the output object of the Nth time downsampling convolution processing is subjected to upsampling convolution processing;

Extracting second feature information, wherein the second feature information is feature information of an output object of the most recent upsampling convolution process;

If the iterative process of extracting the second feature information is not completed N times, based on the currently extracted second feature information, an upsampling convolution process is performed on the output object of the latest upsampling convolution process, and then the extracting second Step of characteristic information;

If the iterative process of extracting the second feature information is performed N times, based on the pre-trained classifier and the second feature information extracted at the Nth time, plaque segmentation is performed on the output object of the latest upsampling convolution processing in order to Determining whether a plaque exists in the blood vessel wall image based on the segmentation result;

Wherein, the N is not less than 2.

A second aspect of the present application provides a blood vessel wall plaque segmentation device, including:

An acquisition unit, a first feature extraction unit, a down-sampled convolution processing unit, a second feature extraction unit, an up-sampled convolution processing unit, and a segmentation unit;

The acquiring unit is configured to acquire an image of a blood vessel wall;

The down-sampling convolution processing unit is configured to trigger the first feature extraction unit after performing down-sampling convolution processing on the blood vessel wall image; when the iterative process of extracting the first feature information is not completed N times, based on the current The first feature information extracted by the first feature extraction unit performs down-sampling convolution processing on the latest output object of the down-sampling convolution processing unit, and then triggers the first feature extraction unit;

The first feature extraction unit is configured to extract first feature information, where the first feature information is feature information of an output object of a recent down-sampling convolution process;

The up-sampling convolution processing unit is configured to: when completing the N iterative process of extracting the first feature information, perform convolution on the down-sampling based on the first feature information extracted by the first feature extraction unit for the Nth time. The object outputted by the processing unit for the Nth time is subjected to upsampling convolution processing, and then the second feature extraction unit is triggered; when the iterative process of extracting the second feature information for N times is not completed, the current feature extraction unit is currently used The obtained second feature information performs upsampling convolution processing on the object output by the upsampling convolution processing unit last time, and then triggers the second feature extraction unit;

The second feature extraction unit is configured to: extract second feature information, wherein the second feature information is feature information of a last output object of the upsampling convolution processing unit;

The segmentation unit is used to perform speckle on the output object of the most recent upsampling convolution processing based on the pre-trained classifier and the second feature information extracted at the N iteration process of extracting the second feature information. Block segmentation, so as to determine whether a plaque exists in the blood vessel wall image based on the segmentation result;

Wherein, the N is not less than 2.

A third aspect of the present application provides a blood vessel wall plaque segmentation device including a memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor executes the computer program At this time, the above-mentioned method for segmenting a blood vessel wall plaque provided by the first aspect of the present application is implemented.

A fourth aspect of the present application provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the blood vessel wall plaque segmentation method provided by the first aspect of the present application is implemented.

As can be seen from the above, on the one hand, the scheme of the present application realizes automatic segmentation of plaques in a blood vessel wall image by extracting feature information (such as second feature information) of the blood vessel wall image and inputting a pre-trained classifier for recognition. Since the plaque segmentation is performed on the vessel wall image automatically by the machine, it is easy to know whether there is a plaque in the vessel wall image through the segmentation result. Compared with the traditional manual judgment method by a medical staff or a medical expert, the solution of this application It can effectively improve the recognition efficiency of blood vessel wall plaques. On the other hand, since the second feature information of the input classifier in the solution of this application is obtained through multiple feature extraction, down-sampling convolution processing and up-sampling convolution processing, The second feature information can better characterize deeper features in the vessel wall image, so that the image segmentation result based on the second feature information is more accurate.

BRIEF DESCRIPTION OF THE DRAWINGS

1-a is a schematic flowchart of an embodiment of a method for segmenting a blood vessel wall plaque provided by this application;

FIG. 1-b is a schematic diagram of a Dense network structure provided by this application;

FIG. 2 is a schematic diagram of a network structure for implementing a plaque segmentation method of a blood vessel wall in an application scenario provided by the present application; FIG.

3 is a schematic structural diagram of an embodiment of a blood vessel wall plaque segmentation device provided by the present application;

FIG. 4 is a schematic structural diagram of another embodiment of a blood vessel wall plaque segmentation device provided by the present application.

detailed description

In order to make the object, features, and advantages of the present application more obvious and easier to understand, the technical solutions in the embodiments of the present application will be clearly and completely described with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described The embodiments are only a part of the embodiments of this application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without making creative labor fall into the protection scope of the present application.

As shown in FIG. 1-a, a method for segmenting a blood vessel wall plaque in the embodiment of the present application includes:

Step 101: Obtain a blood vessel wall image;

In an application scenario, step 101 may be represented as: acquiring a blood vessel wall image by using Magnetic Resonance Imaging (MRI) technology, and the blood vessel wall image obtained at this time is an MRI image. The following explains the MRI technology: MRI technology is a technology that obtains an image of the internal structure of the human body through a magnetic field. It has the advantage of non-trauma, so the patient can be well protected during the examination. In this embodiment, the blood vessel wall images of the human cervical artery can be acquired by MRI technology. Of course, the blood vessel wall images of other blood vessels of the human body can also be acquired by MRI technology, which is not limited here.

In another application scenario, an image of a blood vessel wall may also be obtained by an ultrasonic diagnostic method (for example, B-ultrasound). Alternatively, a blood vessel wall image to be identified may also be acquired (for example, imported) from an existing blood vessel wall image database, which is not limited herein.

Step 102: Perform down-sampling convolution processing on the blood vessel wall image;

In step 102, the above-mentioned blood vessel wall image (the original blood vessel wall image or the normalized blood vessel wall image) is subjected to down-sampling convolution processing.

Specifically, step 102 includes: extracting feature information in the blood vessel wall image, and performing down-sampling convolution processing on the blood vessel wall image based on the extracted feature information.

Optionally, the feature information in the blood vessel wall image is extracted through a convolution process, and the formula applied to the convolution process can be expressed as:

Among them, i, j are the pixel positions of the image, I, K respectively represent the image and the convolution kernel, and m, n are the width and height of the convolution kernel, respectively.

Alternatively, the feature information in the blood vessel wall image may be extracted based on a Dense network or other image feature extraction algorithms, which is not limited herein.

Because the format of the obtained blood vessel wall images may not be the same (for example, there are two different formats of blood vessel wall images obtained based on MRI technology: 176 * 132, 132 * 176), in order to facilitate the training of the network and reduce redundant parameters, After obtaining the blood vessel wall image, the size of the obtained blood vessel wall image is normalized, so that the size of the normalized blood vessel wall image is a uniform size. Optionally, after step 101 and before step 102, the method may further include: performing normalization processing on the image size of the obtained blood vessel wall image to obtain a blood vessel wall image of a preset size. Then, step 102 may be specifically embodied as: performing down-sampling convolution processing on the blood vessel wall image of the preset size.

Specifically, the preset size can be set to 128 * 128, for example. Of course, the size can also be preset to other sizes, which is not limited here.

Step 103: Extract first feature information;

The first feature information is the feature information of the output object of the latest down-sampling convolution process.

In the embodiment of the present application, the feature information (that is, the first feature information) in the output object of the latest down-sampling convolution processing may be extracted based on the image feature extraction technology.

Optionally, in the embodiment of the present application, the first feature information is extracted based on the Dense network. Specifically, the structure diagram of the Dense network can be shown in Figure 1-b. The Dense network includes a convolution layer and an Elu nonlinear activation function. After the input object is processed by the convolution layer and the Elu nonlinear function (the processing is performed in Figure 1-b is expressed as "Convolution + Elu"). The obtained output needs to be superimposed with the input, that is, the input of each layer comes from the output of all the previous layers. If this process is expressed by a function, it can be expressed as: x _l = H _l ([x ₀ , x ₁ , ..., x _l-1 ]). Among them, [x ₀ , x ₁ , ..., x _l-1 ] represents the superposition of the output of layers 0 to l-1, and H _l represents a non-linear transformation. By using the Dense network, the transfer of features is enhanced, the feature information can be used more effectively, the disappearance of gradients is reduced, and the number of parameters is reduced to a certain extent. Dense network input and output pictures have the same size. The Dense network uses the Elu non-linear activation function. The positive characteristics of the Elu non-linear activation function alleviate the problem of gradient disappearance. Compared with the traditional relu activation function, the negative value reduces the computational complexity and meets the requirements of zero mean. Calculation deviation.

Of course, in step 103, the first feature information may also be extracted based on other neural networks or image feature extraction algorithms, which is not limited here.

Step 104: If the iterative process of extracting the first feature information is not completed N times, perform down-sampling convolution processing on the output object of the latest down-sampling convolution processing based on the currently extracted first feature information;

In step 104, based on the first feature information extracted in step 103, downsampling convolution processing is performed on the output object of the latest downsampling convolution processing, and the resolution of the image is reduced so that deeper layers in the image can be extracted in subsequent steps. The characteristic information.

In step 104, the output object of the latest down-sampling convolution process and the currently extracted first feature information may be input to a down-sampling layer (which can be understood as a pooling layer) for down-sampling convolution processing, and the output of the down-sampling layer It is the output object of the down-sampled convolution processing.

After performing down-sampling convolution processing on the output object of the latest down-sampling convolution processing based on the currently extracted first feature information, return to step 103 to perform step 103 iteratively. Through this iterative process, deep feature information in the vessel wall image can be gradually extracted.

Wherein, the N is a preset value not less than 2. Optionally, N is 4.

Step 105: If the iterative process of extracting the first feature information is completed N times, based on the first feature information extracted the Nth time, perform an upsampling convolution process on the output object of the latest downsampling convolution process;

Because in the iterative process of extracting the first feature information, the image is compressed after being down-sampled and convolved. Therefore, in the embodiment of the present application, after completing the N iterative process of extracting the first feature information, the compressed The image is restored. The process of this restoration can be understood as the reverse operation of the aforementioned compression process.

Specifically, when the iterative process of extracting the first feature information is performed N times, based on the first feature information extracted at the Nth time, the output object of the latest (i.e., Nth) downsampling convolution processing is up-sampled and rolled. Product processing to gradually restore the resolution of the image.

In step 105, the output object of the latest down-sampling convolution processing and the N-th extracted first feature information may be input to an up-sampling layer for up-sampling convolution processing, and the output of the up-sampling layer is the current upsampling Output object for convolution processing.

Step 106: Extract second feature information;

The second feature information is the feature information of the output object of the latest up-sampling convolution process.

Optionally, in the embodiment of the present application, the second feature information is extracted based on the Dense network. Specifically, for a description of the Dense network, reference may be made to the description in step 103, and details are not described herein again.

Of course, in step 106, the second feature information may also be extracted based on other neural networks or image feature extraction algorithms, which is not limited here.

Step 107: If the iterative process of extracting the second feature information is not completed N times, perform upsampling convolution processing on the output object of the latest upsampling convolution processing based on the currently extracted second feature information, and then return to step 106;

In the embodiment of the present application, when the iterative process of extracting the second feature information is not completed N times (abbreviated as the incomplete iterative process in FIG. 1-a), it indicates that the currently compressed image of the blood vessel wall still needs to be restored. Step 107. Through this iterative process, the blood vessel wall image can be restored step by step.

In step 107, the currently extracted second feature information and the output object of the latest upsampling convolution processing may be input to the upsampling layer for upsampling convolution processing, and the output of the upsampling layer is the current upsampling volume. The output object of the product processing.

Step 108: If the iterative process of extracting the second feature information is completed N times, based on the pre-trained classifier and the second feature information extracted the Nth time, perform plaque extraction on the output object of the latest upsampling convolution processing. Segmentation to determine whether there are plaques in the blood vessel wall image based on the segmentation results;

When the iterative process of extracting the second feature information is completed N times, it indicates that the current feature extraction process for the above-mentioned blood vessel wall image has been completed. (That is, the image of the restored blood vessel wall) input a pre-trained classifier (such as a softmax classifier) to perform plaque segmentation, that is, to isolate the plaque and background information in the output object, so that the segmentation result can be used to determine the Whether there are plaques in the vessel wall image. Specifically, based on the second feature information extracted at the Nth time, the output object of the last upsampling convolution process, and the above-mentioned classifier, each pixel in the output object may be classified as foreground information (such as plaque) or background information. , So as to separate the plaque and background in the image of the blood vessel wall.

It should be noted that, for the network for automatically segmenting the blood vessel wall image in the embodiments of the present application (hereinafter collectively referred to as a segmentation network, for example, the segmentation network may be the aforementioned Dense network, downsampling layer, upsampling layer, and classification Device, etc.), can be obtained through pre-training. In practical applications, the segmentation network can be trained by acquiring multiple vessel wall images used to train the segmentation network, and the segmentation network can be optimized based on Adam's optimization algorithm. Specifically, the process of optimizing the segmented network based on the Adam optimization algorithm can be implemented by referring to the existing technology, and is not repeated here.

As mentioned earlier, in step 103 and step 106, the first feature information and the second feature information may be extracted based on the Dense network. In this application scenario, the following constraints can be set: 1. In the above iterative process of extracting the first feature information, the number of convolution kernels of the Dense network used to extract the first feature information for the n + 1th time is the first Double the number of convolution kernels of the Dense network used to extract the first feature information n times; and in the above iterative process of extracting the second feature information, the Dense used to extract the second feature information n + 1 times The number of convolution kernels of the network is one half of the number of convolution kernels of the Dense network used for extracting the second feature information for the nth time; and the volume of the Dense network used for extracting the first feature information for the nth time. The number of kernels is equal to the number of kernels of the Dense network used for the first extraction of the second feature information, where n ∈ [1, N).

As can be seen from the above, on the one hand, the vascular wall plaque segmentation method in the embodiment of the present application realizes the identification of blood vessels by extracting feature information (such as second feature information) of the blood vessel wall image and inputting a pre-trained classifier for recognition. The automatic segmentation of plaque in the wall image is because the plaque segmentation of the vascular wall image is automatically performed by the machine. Therefore, it is easy to know whether there is a plaque in the vascular wall image through the segmentation result. Compared with the traditional medical staff or medicine, The method of manual judgment by experts can effectively improve the recognition efficiency of vascular wall plaques. On the other hand, because the second feature information of the input classifier in the scheme of this application is obtained through multiple feature extraction and down-sampling convolution processing. And the upsampling convolution processing is obtained, therefore, the second feature information can better characterize deeper features in the vessel wall image, thereby making the image segmentation result based on the second feature information more accurate.

To facilitate a better understanding of the vascular wall plaque segmentation method in the embodiment shown in FIG. 1-a, the above vascular wall plaque segmentation method will be described in a specific application scenario below. The schematic diagram of the network structure in this application scenario can be shown in Figure 2. As can be seen from Figure 2, the segmentation network in this application scenario includes two parts: compression and extraction features and decompression image restoration. The two parts are completely symmetrical to ensure The divided image is the same size as the original image. Vessel wall images are processed by Dense network (refer to the structure and related description of Dense network in Figure 1-b) and then input compression and extraction feature parts for processing.

It can be seen from Figure 2 that the compression extraction feature part and the decompressed image restoration part both include four segments of processing. For the compression extraction feature part, each segment is composed of the downsampling layer and the Dense network (refer to the structure and related description of the Dense network in Figure 1-b). ) To gradually extract deeper feature information of the image. Similarly, the recovery part of the decompressed image also includes four segments of processing (that is, the aforementioned N is taken as 4), and each segment is composed of an upsampling layer and a Dense network to gradually restore the image.

Optionally, in the compression extraction feature part, the size of the convolution kernel of the Dense network used in each processing is 5 * 5, 5 * 5, 5 * 5, and 5 * 5, respectively. The number of convolution kernels is 32, 64, 128, and 256. The size of the first input Dense network in the compression extraction feature is 128 * 128, and the size of the image output by compression extraction is 8 * 8.

Correspondingly, in the decompressed image recovery part, the size of the convolution kernel of the Dense network used for each segment of processing is 5 * 5, 5 * 5, 5 * 5, and 5 * 5; the size of the Dense network used for each segment of processing The number of convolution kernels are 256, 128, 64, and 32 respectively. The size of the first input image to the Dense network is 128 * 128, and the size of the image output from the compression and extraction feature part is 8 * 8. The size of the image input to the Dense network for the first time in the decompressed image recovery part is 8 * 8, and the size of the image output from the compression extraction feature part is 128 * 128.

After the decompressed image restoration part is processed, the output of the decompressed image restoration part is input to the softmax classifier, and the softmax classifier performs plaque segmentation on the vascular wall image, that is, the plaque in the image is separated from the background (that is, Output segmentation result) to determine whether a plaque exists in the blood vessel wall image based on the segmentation result.

FIG. 3 provides a blood vessel wall plaque segmentation device according to an embodiment of the present application. As shown in FIG. 3, the vascular wall plaque segmentation device mainly includes an acquisition unit 301, a first feature extraction unit 302, a down-sampling convolution processing unit 303, a second feature extraction unit 304, an up-sampling convolution processing unit 305, and Divide unit 306.

The obtaining unit 301 is configured to: obtain a blood vessel wall image;

The down-sampling convolution processing unit 303 is configured to trigger the first feature extraction unit 302 after performing down-sampling convolution processing on the blood vessel wall image obtained by the obtaining unit 301; when the iterative process of extracting the first feature information is not completed N times, Performing downsampling convolution processing on the latest output object of the downsampling convolution processing unit 303 based on the first feature information extracted by the current first feature extraction unit 302, and then triggering the first feature extraction unit 302;

The first feature extraction unit 302 is configured to extract first feature information, where the first feature information is feature information of an output object of a recent down-sampling convolution process;

The up-sampling convolution processing unit 305 is configured to complete the N-th iteration process of extracting the first feature information based on the first feature information extracted by the first feature extraction unit 302 for the Nth time, and perform the N-th The current output object is subjected to up-sampling convolution processing, and then the second feature extraction unit 304 is triggered; when the iterative process of extracting the second feature information is not completed N times, based on the The second feature information performs upsampling convolution processing on the object that was last output by the upsampling convolution processing unit 305, and then triggers the second feature extraction unit 304;

The second feature extraction unit 304 is configured to extract second feature information, where the second feature information is feature information of the latest output object of the upsampling convolution processing unit 305;

The segmentation unit 306 is configured to: when completing the iterative process of extracting the second feature information N times, based on the pre-trained classifier and the second feature information extracted the Nth time, perform an output object on the latest upsampling convolution processing Plaque segmentation, so as to determine whether a plaque exists in the blood vessel wall image based on the segmentation result;

Wherein, the aforementioned N is not less than 2, preferably, N is taken as 4.

Optionally, the first feature extraction unit 302 is specifically configured to: extract the first feature information based on the Dense network; the second feature extraction unit 304 is specifically configured to: extract the second feature information based on the Dense network.

Optionally, for the above-mentioned blood vessel wall image, the number of convolution kernels of the Dense network used by the first feature extraction unit 302 to extract the first feature information for the n + 1th time is the Dense used to extract the first feature information for the nth time. The number of convolution kernels of the network is doubled; and the number of convolution kernels of the Dense network used by the second feature extraction unit 304 to extract the second feature information for the n + 1th time is to extract the second feature information for the nth time. One-half the number of convolution kernels of the Dense network used; and, the number of convolution kernels of the Dense network used by the first feature extraction unit 302 to extract the first feature information n times is equal to the second feature extraction unit 304 The number of convolution kernels of the Dense network used for the first extraction of the second feature information; where n ∈ [1, N).

Optionally, the blood vessel wall plaque segmentation device further includes a normalization unit configured to perform normalization processing on the image size of the blood vessel wall image acquired by the obtaining unit 301 to obtain a blood vessel wall image of a preset size. The down-sampling convolution processing unit 303 is specifically configured to perform down-sampling convolution processing on the blood vessel wall image obtained by the normalizing unit.

It should be noted that the vascular wall plaque segmentation device can be used to implement the vascular wall plaque segmentation method provided by the foregoing method embodiment. In the vascular wall plaque segmentation device illustrated in FIG. 3, the division of each functional module is merely an example. In actual applications, the above functions may be allocated by the needs of, for example, the configuration requirements of the corresponding hardware or the convenience of software implementation. Different functional modules are completed, that is, the internal structure of the vascular wall plaque segmentation device is divided into different functional modules to complete all or part of the functions described above. Moreover, in actual applications, the corresponding functional modules in this embodiment may be implemented by corresponding hardware, or may be completed by corresponding hardware executing corresponding software. The embodiments described in this specification can apply the above-mentioned description principles, which will not be described in detail below.

As can be seen from the above, in the embodiment of the present application, feature information (such as the second feature information) of a blood vessel wall image is extracted and a pre-trained classifier is used for recognition, so as to automatically segment plaques in the blood vessel wall image. The plaque segmentation of the vessel wall image is performed automatically by the machine. Therefore, it is easy to know whether there is a plaque in the vessel wall image through the segmentation result. Compared with the traditional manual judgment method by a medical staff or a medical expert, the solution of the present application can be effective Improve the recognition efficiency of blood vessel wall plaques; on the other hand, since the second feature information of the input classifier in the solution of this application is obtained through multiple feature extraction, down-sampling convolution processing and up-sampling convolution processing, The two feature information can better characterize the deeper features in the vessel wall image, so that the image segmentation result based on the second feature information is more accurate.

An embodiment of the present application provides a vascular wall plaque segmentation device. Referring to FIG. 4, the vascular wall plaque segmentation device includes:

The memory 41, the processor 42, and a computer program stored on the memory 41 and executable on the processor 42. When the processor 42 executes the computer program, the blood vessel wall plaque segmentation method described in the foregoing method embodiment is implemented.

Further, the blood vessel wall plaque segmentation device further includes:

At least one input device 43 and at least one output device 44.

The memory 41, the processor 42, the input device 43, and the output device 44 are connected via a bus 45.

The input device 43 and the output device 44 may be antennas.

The memory 41 may be a high-speed random access memory (RAM, Random Access Memory) memory, or may be a non-volatile memory (non-volatile memory), such as a magnetic disk memory. The memory 41 is configured to store a set of executable program code, and the processor 42 is coupled to the memory 41.

Further, an embodiment of the present application further provides a computer-readable storage medium. The computer-readable storage medium may be a blood vessel wall plaque segmentation device provided in the foregoing embodiments. The computer-readable storage medium may be The memory in the aforementioned embodiment shown in FIG. 4. A computer program is stored on the computer-readable storage medium, and when the program is executed by a processor, the power allocation method described in the foregoing method embodiment is implemented. Further, the computer-storable medium may also be various media that can store program codes, such as a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a RAM, a magnetic disk, or an optical disk.

In the several embodiments provided in this application, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules is only a logical function division. In actual implementation, there may be another division manner. For example, multiple modules or components may be combined or Can be integrated into another system, or some features can be ignored or not implemented. In addition, the displayed or discussed mutual coupling or direct coupling or communication connection may be indirect coupling or communication connection through some interfaces, devices or modules, which may be electrical, mechanical or other forms.

The modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, may be located in one place, or may be distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the objective of the solution of this embodiment.

In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist separately physically, or two or more modules may be integrated into one module. The above integrated modules can be implemented in the form of hardware or software functional modules.

When the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially a part that contributes to the existing technology or all or part of the technical solution can be embodied in the form of a software product, which is stored in a readable storage The medium includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in the embodiments of the present application. The foregoing readable storage medium includes: various media that can store program codes, such as a U disk, a mobile hard disk, a ROM, a RAM, a magnetic disk, or an optical disk.

It should be noted that, for the foregoing method embodiments, for simplicity of description, they are all described as a series of action combinations, but those skilled in the art should know that this application is not limited by the described action order. Because according to the present application, certain steps may be performed in another order or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required for this application.

In the above embodiments, the description of each embodiment has its own emphasis. For a part that is not described in detail in an embodiment, reference may be made to related descriptions in other embodiments.

The above is a description of the vascular wall plaque segmentation method, device and computer-readable storage medium provided by the present application. For those skilled in the art, according to the ideas of the embodiments of the present application, there will be specific implementations and application scopes. In summary, the content of this specification should not be construed as a limitation on this application.

Claims

A plaque segmentation method for a blood vessel wall, comprising:

Obtain a blood vessel wall image;

Performing down-sampling convolution processing on the blood vessel wall image;

Extracting first feature information, where the first feature information is feature information of an output object of a recent down-sampling convolution process;

If the iterative process of extracting the first feature information is not completed N times, the output object of the latest down-sampling convolution process is subjected to down-sampling convolution processing based on the currently extracted first feature information, and then the extracting the first feature is performed iteratively Information steps

If the iterative process of extracting the first feature information is performed N times, based on the first feature information extracted at the Nth time, the output object of the Nth time downsampling convolution processing is subjected to upsampling convolution processing;

Extracting second feature information, wherein the second feature information is feature information of an output object of the most recent upsampling convolution process;

If the iterative process of extracting the second feature information is not completed N times, based on the currently extracted second feature information, an upsampling convolution process is performed on the output object of the latest upsampling convolution process, and then the extracting second Step of characteristic information;

If the iterative process of extracting the second feature information is performed N times, based on the pre-trained classifier and the second feature information extracted at the Nth time, plaque segmentation is performed on the output object of the latest upsampling convolution processing in order to Determining whether a plaque exists in the blood vessel wall image based on the segmentation result;

Wherein, the N is not less than 2.
The method according to claim 1, wherein the extracting the first feature information is: extracting the first feature information based on a Dense network;

The extracting the second feature information is: extracting the second feature information based on the Dense network.
The plaque segmentation method for a blood vessel wall according to claim 2, wherein:

In the iterative process of extracting the first feature information, the number of convolution kernels of the Dense network used to extract the first feature information for the n + 1th time is the volume of the Dense network used to extract the first feature information for the nth time. Double the number of cores;

Moreover, in the iterative process of extracting the second feature information, the number of convolution kernels of the Dense network used to extract the second feature information for the n + 1th time is the Dense network used to extract the second feature information for the nth time. One-half the number of convolution kernels;

In addition, the number of convolution kernels of the Dense network used to extract the first feature information for the nth time is equal to the number of convolution kernels of the Dense network used to extract the second feature information for the first time;

Among them, n ∈ [1, N).
The method for segmenting a blood vessel wall plaque according to any one of claims 1 to 3, wherein after acquiring the blood vessel wall image, the method further comprises:

Normalize the image size of the obtained blood vessel wall image to obtain a blood vessel wall image of a preset size;

The down-sampling convolution processing on the blood vessel wall image is:

Performing down-sampling convolution processing on the blood vessel wall image of the preset size.
The method for segmenting a blood vessel wall plaque according to any one of claims 1 to 3, wherein the N is 4.
A blood vessel wall plaque segmentation device, comprising: an acquisition unit, a first feature extraction unit, a down-sampling convolution processing unit, a second feature extraction unit, an up-sampling convolution processing unit, and a segmentation unit;

The acquiring unit is configured to acquire an image of a blood vessel wall;

The down-sampling convolution processing unit is configured to trigger the first feature extraction unit after performing down-sampling convolution processing on the blood vessel wall image; when the iterative process of extracting the first feature information is not completed N times, based on the current The first feature information extracted by the first feature extraction unit performs down-sampling convolution processing on the latest output object of the down-sampling convolution processing unit, and then triggers the first feature extraction unit;

The first feature extraction unit is configured to extract first feature information, where the first feature information is feature information of an output object of a recent down-sampling convolution process;

The up-sampling convolution processing unit is configured to: when completing the N iterative process of extracting the first feature information, perform convolution on the down-sampling based on the first feature information extracted by the first feature extraction unit for the Nth time. The object outputted by the processing unit for the Nth time is subjected to upsampling convolution processing, and then the second feature extraction unit is triggered; when the iterative process of extracting the second feature information for N times is not completed, the current feature extraction unit is currently used The obtained second feature information performs upsampling convolution processing on the object output by the upsampling convolution processing unit last time, and then triggers the second feature extraction unit;

The second feature extraction unit is configured to: extract second feature information, wherein the second feature information is feature information of a last output object of the upsampling convolution processing unit;

The segmentation unit is used to perform speckle on the output object of the most recent upsampling convolution processing based on the pre-trained classifier and the second feature information extracted at the N iteration process of extracting the second feature information. Block segmentation, so as to determine whether a plaque exists in the blood vessel wall image based on the segmentation result;

Wherein, the N is not less than 2.
The vascular plaque segmentation device according to claim 6, wherein the first feature extraction unit is specifically configured to extract first feature information based on a Dense network;

The second feature extraction unit is specifically configured to extract the second feature information based on the Dense network.
The vascular wall plaque segmentation device according to claim 7, wherein for the blood vessel wall image, the first feature extraction unit extracts the first feature information by a convolution of the Dense network for the n + 1th time The number of cores is twice the number of convolution cores of the Dense network used for extracting the first feature information for the nth time;

In addition, the number of convolution kernels of the Dense network used for extracting the second feature information by the n + 1th time is the number of convolution kernels of the Dense network used by the nth extraction of the second feature information. One-half of

Moreover, the number of convolution kernels of the Dense network used by the first feature extraction unit to extract the first feature information for the nth time is equal to Number of convolution kernels;

Among them, n ∈ [1, N).
A blood vessel wall plaque segmentation device, comprising: a memory, a processor, and a computer program stored on the memory and operable on the processor. The processor is implemented when the processor executes the computer program. A method according to any one of claims 1 to 5.
A computer-readable storage medium having stored thereon a computer program, characterized in that when the computer program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.