CN116883358A

CN116883358A - Vascular image processing method and device, electronic equipment and storage medium

Info

Publication number: CN116883358A
Application number: CN202310844700.8A
Authority: CN
Inventors: 刘洵; 陈树湛
Original assignee: Shanghai Bodong Medical Technology Co ltd
Current assignee: Shanghai Bodong Medical Technology Co ltd
Priority date: 2023-07-11
Filing date: 2023-07-11
Publication date: 2023-10-13

Abstract

The invention discloses a blood vessel image processing method, a blood vessel image processing device, electronic equipment and a storage medium. Comprising the following steps: acquiring a plurality of blood vessel scanning data of a blood vessel to be detected in a preset period, and respectively registering and segmenting the plurality of blood vessel scanning data to obtain blood vessel segmentation data in each blood vessel scanning data; extracting blood vessel position points based on each blood vessel segmentation data, and generating displacement field data of the blood vessel to be detected in a preset period based on the position information of the blood vessel position points in each blood vessel scanning data; a spatial movement parameter for each vessel location point is determined based on the displacement field data, and a set of target location points in the vessel is determined based on the spatial movement parameter. According to the invention, registration and segmentation processing are respectively carried out on a plurality of blood vessel scanning data, blood vessel position points are extracted from the processing results, and corresponding space movement parameters and target position point sets are determined through blood vessel position point information, so that the accuracy of a blood vessel image processing technology is improved.

Description

Vascular image processing method and device, electronic equipment and storage medium

Technical Field

The present invention relates to the field of image processing technologies, and in particular, to a blood vessel image processing method, a blood vessel image processing device, an electronic device, and a storage medium.

Background

With the development of image processing technology, image processing technology has been gradually applied to the clinical medical field. Such as image processing techniques, are used in the treatment of heart valve disease.

Currently, a heart valve disease is usually treated by using a transcatheter aortic valve replacement (Transcatheter aortic valve replacement, TAVR) operation, scanning data of an operation site is acquired before the operation by using an electronic computed tomography (Computed Tomography, CT) image technology for reference by an operator, and the operator selects an operation position according to prior experience to complete the operation.

The images obtained through the mode can only roughly observe blood vessels, the value of image reference is low, and the positioning of the operation position has strong dependence on experience of an operator.

Disclosure of Invention

The invention provides a blood vessel image processing method, a blood vessel image processing device, electronic equipment and a storage medium, which are used for providing an operable position of a blood vessel by processing acquired images.

According to an aspect of the present invention, there is provided a blood vessel image processing method including:

acquiring a plurality of blood vessel scanning data of a blood vessel to be detected in a preset period, and respectively registering and segmenting the plurality of blood vessel scanning data to obtain blood vessel segmentation data in each blood vessel scanning data;

Extracting blood vessel position points based on each blood vessel segmentation data, and generating displacement field data of the blood vessel to be detected in a preset period based on the position information of the blood vessel position points in each blood vessel scanning data;

a spatial movement parameter for each vessel location point is determined based on the displacement field data, and a set of target location points in the vessel is determined based on the spatial movement parameter.

Optionally, the preset period is a cardiac period, and the plurality of vessel scan data are multi-phase vessel scan data obtained by performing vessel scan based on a preset time interval in the cardiac period.

Optionally, registering and segmenting the plurality of vessel scan data respectively to obtain vessel segmentation data in each vessel scan data, including:

inputting a plurality of blood vessel scanning data into a pre-trained registration segmentation joint model to obtain initial segmentation data of each blood vessel scanning data and a registration transformation matrix of each blood vessel scanning data relative to reference blood vessel scanning data respectively;

and respectively carrying out conversion processing on the corresponding initial segmentation data based on each registration transformation matrix to obtain vessel segmentation data of each vessel scanning data.

Optionally, the training method for registering the segmentation joint model includes:

Acquiring a plurality of groups of blood vessel sample data, wherein any group of blood vessel sample data is a plurality of blood vessel sample data in any preset period;

the following iterative training steps are executed based on the multiple sets of blood vessel sample data until the iterative stopping condition is met, so that a registration segmentation joint model with completed training is obtained:

inputting any group of blood vessel sample data into a registration segmentation joint model to be trained to obtain blood vessel segmentation data and registration results respectively corresponding to the blood vessel sample data;

and determining a joint model loss based on the label information corresponding to the blood vessel sample data, the blood vessel segmentation data corresponding to the blood vessel sample data and the registration result, and adjusting parameters of the registration segmentation joint model based on the joint model loss.

Optionally, the joint model loss is determined based on the registration loss and the segmentation loss;

any group of blood vessel sample data comprises reference blood vessel data and motion blood vessel data;

the segmentation loss item is determined based on a first loss item formed by the blood vessel segmentation data corresponding to the reference blood vessel data and the label information corresponding to the reference blood vessel data and/or a second loss item formed by the blood vessel segmentation data corresponding to the motion blood vessel data and the label information corresponding to the motion blood vessel data;

The registration loss term is based on any of the following: performing registration processing on the tag information of the moving blood vessel data based on a registration result corresponding to the moving blood vessel data to obtain registration tag information, and generating a registration loss item based on the registration tag information and the tag information of the reference blood vessel data;

performing registration processing on the vessel segmentation data based on registration results corresponding to the moving vessel data to obtain registration segmentation results of the moving vessel data, and generating registration loss items based on the registration segmentation results and label information of the reference vessel data;

and carrying out registration processing on the label information of the moving blood vessel data based on a registration result corresponding to the moving blood vessel data to obtain registration label information, and generating a registration loss item based on the registration label information and blood vessel segmentation data of the reference blood vessel data.

Optionally, extracting the vessel location point based on each vessel segmentation data includes:

for any vessel segmentation data, determining a vessel center line of the current vessel segmentation data and each center point on the vessel center line, and determining contour point cloud data of a vessel section contour in the current vessel segmentation data based on each center point, wherein the contour point cloud data is used as a vessel position point.

Optionally, determining contour point cloud data of the vessel cross-section contour in the current vessel segmentation data based on each center point includes:

for any central point, performing expansion processing on the current central point to obtain a blood vessel section profile corresponding to the current central point;

and sampling the blood vessel section profile to obtain profile point data of profile points on the blood vessel section profile.

Optionally, generating displacement field data of the blood vessel to be detected in a preset period based on the position information of the blood vessel position point in each blood vessel scanning data includes:

for any vessel position point, determining the position information of the vessel position point in each vessel segmentation data, and generating displacement field data of the vessel to be detected in a preset period based on the phase sequence of a plurality of vessel segmentation data and the position information of the vessel position point in each vessel segmentation data.

Optionally, the spatial movement parameter includes one or more of an average displacement, a spatial gradient of displacement, and a displacement concussion index;

determining a spatial movement parameter for each vessel location point based on the displacement field data, including one or more of:

for any vessel position point, determining average displacement based on the absolute value average value of a plurality of displacement data of the vessel position point in a preset period;

Determining a displacement spatial gradient based on the displacement change rate of the vascular position points in a preset period;

and determining a displacement concussion index based on the ratio of the displacement variation of the vascular position point in the preset period to the total moving path in the preset period.

Optionally, determining the set of target location points in the vessel based on the spatial movement parameter includes:

for any blood vessel position point, comparing one or more of average displacement, displacement space gradient and displacement concussion index of the blood vessel position point with corresponding threshold values respectively to obtain comparison results;

and screening blood vessel position points meeting the motion stability condition according to the comparison result to form a target position point set.

According to another aspect of the present invention, there is provided a blood vessel image processing apparatus including:

the blood vessel segmentation data acquisition module is used for acquiring a plurality of blood vessel scanning data of a blood vessel to be detected in a preset period, and registering and segmenting the plurality of blood vessel scanning data respectively to obtain blood vessel segmentation data in each blood vessel scanning data;

the displacement field data determining module is used for extracting blood vessel position points based on the blood vessel segmentation data and generating displacement field data of the blood vessel to be detected in a preset period based on the position information of the blood vessel position points in the blood vessel scanning data;

And the target position point set determining module is used for determining the space movement parameter of each blood vessel position point based on the displacement field data and determining the target position point set in the blood vessel based on the space movement parameter.

According to another aspect of the present invention, there is provided an electronic apparatus including:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein,,

the memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the vascular image processing method of any of the embodiments of the present invention.

According to another aspect of the present invention, there is provided a computer readable storage medium storing computer instructions for causing a processor to execute a blood vessel image processing method of any one of the embodiments of the present invention.

According to the technical scheme provided by the embodiment of the invention, the registration and segmentation processing are respectively carried out on the plurality of blood vessel scanning data, the blood vessel position points are extracted from the processing results, and the corresponding space movement parameters and the target position point set are determined through the blood vessel position point information, so that the problem that the movement condition of the blood vessel cannot be accurately determined is solved, the space movement condition of the position points on the blood vessel in the cardiac cycle can be more accurately determined, and the accuracy of a blood vessel image processing technology is improved.

It should be understood that the description in this section is not intended to identify key or critical features of the embodiments of the application or to delineate the scope of the application. Other features of the present application will become apparent from the description that follows.

The technical scheme of the application obtains, stores and/or processes the data, and accords with the relevant regulations of national laws and regulations.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for the description of the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present application, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a flowchart of a blood vessel image processing method according to an embodiment of the present application;

fig. 2 is a flowchart of a blood vessel image processing method according to a second embodiment of the present application;

fig. 3 is a schematic structural diagram of a blood vessel image processing device according to a third embodiment of the present application;

fig. 4 is a schematic structural diagram of an electronic device implementing a blood vessel image processing method according to an embodiment of the present application.

Detailed Description

In order that those skilled in the art will better understand the present invention, a technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, shall fall within the scope of the present invention.

It should be noted that the terms "first," "second," and the like in the description and the claims of the present invention and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the invention described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

Example 1

Fig. 1 is a flowchart of a blood vessel image processing method according to a first embodiment of the present invention, where the method may be applied to a blood vessel image processing apparatus, and the blood vessel image processing apparatus may be implemented in hardware and/or software, and the blood vessel image processing apparatus may be configured in an electronic device such as a computer, an image processing device, or the like. As shown in fig. 1, the method includes:

s110, acquiring a plurality of blood vessel scanning data of a blood vessel to be detected in a preset period, and respectively registering and segmenting the plurality of blood vessel scanning data to obtain blood vessel segmentation data in each blood vessel scanning data.

The blood vessel to be detected is specifically understood to be a blood vessel of an operation region to be operated by a target object, which may be a human body or an animal body, etc., and the blood vessel scanning data is acquired and processed under the condition that the target object or an associated object of the target object is authorized. The vessel scan data may be specifically understood as medical image data, and may be obtained by scanning a vessel to be detected by angiography (Computed Tomographic Arteriography, CTA), magnetic resonance imaging (Magnetic Resonance Imaging, MRI) or the like. Generally, in order to accurately recognize the shape, size, movement and other conditions of a blood vessel to be detected, a plurality of blood vessel scan data needs to be acquired, a certain preset period may be set through actual needs, and equal time interval sampling is performed on the blood vessel to be detected in the preset period, so as to acquire the plurality of blood vessel scan data in the preset period.

Specifically, the blood vessel to be detected can be scanned by a CT angiography technology, and the blood vessel to be detected is sampled at equal time intervals within a preset period, so that a plurality of blood vessel scanning images, namely a plurality of blood vessel scanning data, are obtained. Optionally, the preset period is a cardiac period, and the plurality of vessel scan data are multi-phase vessel scan data obtained by performing vessel scan based on a preset time interval in the cardiac period.

The cardiac cycle is understood to mean, in particular, a complete cardiac cycle, on an electrocardiogram, the time interval between two R waves. The multi-phase vessel scan data may be specifically understood as vessel scan data of a plurality of time points obtained by scanning a vessel to be detected at equal time intervals in a cardiac cycle, for example, 10% of the cardiac cycle is taken as a time interval, and the vessel is scanned at time points such as 0%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80% and 90% of the cardiac cycle to obtain a plurality of vessel scan data of corresponding time points, where the vessel scan data obtained by scanning at each time point is one phase vessel scan data.

Specifically, the CT angiography technology is used to acquire the vascular scan data in one cardiac cycle according to the time interval, acquire the vascular scan data in multiple phases, and acquire the vascular scan data in one cardiac cycle according to the time interval of 10% of the cardiac cycle, so as to obtain the vascular scan data in 10 phases, and it should be noted that each time the heart contracts and expands to form one cardiac cycle, so that the end diastole can be used as an initial value of the cardiac cycle, the vascular scan data acquired in the end diastole is used as the vascular scan data in 1 phase, and the vascular scan data acquired in sequence later is labeled and sequentially referred to as the vascular scan data in 2 phase, the vascular scan data in 3 phase, the vascular scan data … …, the vascular scan data in 10 phase, and the like. It will be appreciated that the duration of the cardiac cycle varies from one target object to another, and that for each target object the duration of the cardiac cycle of that target object may be detected and the time interval for each data scan determined from the duration.

In this embodiment, a plurality of blood vessel scan data in a preset period are acquired according to an equal time interval for a blood vessel to be detected, and the plurality of blood vessel scan data are respectively registered and segmented to obtain blood vessel segmentation data in the blood vessel scan data, so that the problem that a blood vessel segmentation result is difficult to accurately obtain due to a single-phase blood vessel scan image can be avoided by using the method for determining the blood vessel segmentation data based on the multi-phase blood vessel scan data, and the accuracy of the segmentation result of the blood vessel scan data is improved.

Optionally, registering and segmenting the plurality of vessel scan data by a preset registration segmentation model or registration segmentation algorithm to obtain vessel segmentation images in each vessel scan data. The manner of registration and segmentation is not limited herein. The influence caused by vascular motion is reduced by registering a plurality of vascular scan data in a preset period, and the interference of other tissue data in the vascular scan data on the vascular data processing process is eliminated by dividing the vascular scan data, so that the subsequent processing of the vascular data is facilitated.

S120, extracting blood vessel position points based on the blood vessel segmentation data, and generating displacement field data of the blood vessel to be detected in a preset period based on the position information of the blood vessel position points in the blood vessel scanning data.

The vessel position point is specifically understood to be a point on a vessel contour in the vessel segmentation data, and may be marked by a contour point marking model, or a plurality of arbitrary points on the vessel may be randomly selected as vessel position points, which is not limited herein. The displacement field data can be specifically understood as data representing the position coordinates, displacement and displacement variation of each vascular position point in a preset period, and can be calculated through the position information of each position point in different vascular scanning data. The position information of each blood vessel position point in the corresponding blood vessel segmentation data may be represented by three-dimensional coordinates and/or two-dimensional coordinates, and is not limited herein.

Specifically, the vascular position points can be extracted from each vascular segmentation data through a vascular position point extraction model, meanwhile, the position information of each vascular position point in the vascular scanning data is marked, and the spatial position coordinates and the radial position coordinates corresponding to the vascular position points can be respectively represented by adopting three-dimensional coordinates and two-dimensional coordinates. And generating displacement field data of the blood vessel to be detected in a preset period according to the position coordinate information.

Optionally, extracting the vessel location point based on each vessel segmentation data includes: for any vessel segmentation data, determining a vessel center line of the current vessel segmentation data and each center point on the vessel center line, and determining contour point cloud data of a vessel section contour in the current vessel segmentation data based on each center point, wherein the contour point cloud data is used as a vessel position point.

The central line of the blood vessel can be specifically understood as a line formed by connecting central points of a plurality of cross sections of the blood vessel, the central line can be obtained by carrying out refinement treatment on a blood vessel refinement algorithm in blood vessel segmentation data, the cross sections of the plurality of blood vessels can be obtained by a blood vessel cutting method, the central points are determined according to a cross section central point calculation algorithm, and the central points are connected to obtain the central line. And are not limited herein. The contour point cloud data is specifically understood to be data composed of points on the contour of the blood vessel cross section where the center point on the blood vessel center line corresponding to the blood vessel segmentation data of the different-phase blood vessel scan data is located.

Specifically, the central line of the blood vessel segmentation data can be obtained through a blood vessel thinning algorithm, the points on the central line are all central points of the blood vessel, a plurality of central points on the central line of the blood vessel are obtained through an equidistant sampling method, the cross sectional area of each central point can be obtained through a straightening curve reconstruction algorithm, then the contour points are marked on the contour of the cross section, and the contour point cloud data, namely the position points of the blood vessel, are determined. Illustratively, equidistant resampling is a fixed number M, such as 1024 center points; for each center point, acquiring the cross section of the current center point based on a straightening curve reconstruction algorithm, acquiring contour point coordinates according to a segmentation result, and resampling the contour point coordinates at equal intervals to a fixed number N, such as 512 contour points, so as to generate point cloud data which can be used for indicating the boundary of a blood vessel and is distributed along a central line in multiple layers; the center line is determined and contour point cloud data is generated in the same way in other phase, and it is understood that the number of points in each phase data is the same and is 1024×512.

Optionally, determining contour point cloud data of the vessel cross-section contour in the current vessel segmentation data based on each center point includes: for any central point, performing expansion processing on the current central point to obtain a blood vessel section profile corresponding to the current central point; and sampling the blood vessel section profile to obtain profile point data of profile points on the blood vessel section profile.

Specifically, for any central point on a central line, an image vertical expansion method is adopted to expand the current central point to obtain the profile of the cross section of the blood vessel corresponding to the central point, namely the profile of the cross section of the blood vessel, and the profile point of the cross section of the blood vessel is determined by carrying out point sampling on the profile of the cross section of the blood vessel according to an equidistant resampling method, so that profile point data of the profile point on the profile of the cross section of the blood vessel is obtained.

Optionally, generating displacement field data of the blood vessel to be detected in a preset period based on the position information of the blood vessel position point in each blood vessel scanning data includes: for any vessel position point, determining the position information of the vessel position point in each vessel segmentation data, and generating displacement field data of the vessel to be detected in a preset period based on the phase sequence of a plurality of vessel segmentation data and the position information of the vessel position point in each vessel segmentation data.

Specifically, for any vessel position point, the position information of the vessel position point in each vessel segmentation image needs to be determined, the vessel position point, the position point coordinate information and the phase sequence where the vessel position point and the position point coordinate information are located are taken together as the displacement data of the vessel position point, and the displacement data of the vessel position point in all phase vessel scanning data are combined to generate the displacement field data of the vessel to be detected in a preset period.

In this embodiment, by extracting the vascular position points from each vascular segmentation data and introducing a calculation method of the contour point cloud, the position information of the vascular position points in each vascular scanning data is determined, so as to generate displacement field data of the blood vessel to be detected in a preset period, and a data basis is provided for the subsequent calculation of the spatial movement parameters of each vascular position point and the target position point set, so that the knowledge of the displacement condition of the blood vessel in the cardiac period is facilitated.

S130, determining a space movement parameter of each blood vessel position point based on the displacement field data, and determining a target position point set in the blood vessel based on the space movement parameter.

The target position point set may be understood as a set of vessel position points satisfying a motion condition, for example, a set of vessel position points with relatively stable motion, where the target position point set may determine the motion condition according to a vessel processing requirement. The target position point set may be determined by a change condition of a spatial movement parameter, for example, a certain change threshold may be set, and if the spatial movement parameter value of the blood vessel position point is smaller than the set change threshold, it may be determined that the current blood vessel position point is relatively stable, and it may be regarded as the target position point.

Specifically, by performing displacement algorithm processing on displacement field data, calculating spatial movement parameter values of each vascular position point, wherein the spatial movement parameters include, but are not limited to, spatial displacement, average displacement, spatial displacement gradient and the like, determining a target position point set by setting a threshold value, and comparing the spatial movement parameter values with the set threshold value, and meeting the condition that the spatial movement parameter values are smaller than the set threshold value, then the corresponding vascular position point is the target position point, and determining all vascular position points meeting the condition by the method to obtain the target position point set in the blood vessel.

Optionally, the spatial movement parameter includes one or more of an average displacement, a spatial gradient of displacement, and a displacement concussion index; determining a spatial movement parameter for each vessel location point based on the displacement field data, including one or more of: for any vessel position point, determining average displacement based on the absolute value average value of a plurality of displacement data of the vessel position point in a preset period; determining a displacement spatial gradient based on the displacement change rate of the vascular position points in a preset period; and determining a displacement concussion index based on the ratio of the displacement variation of the vascular position point in the preset period to the total moving path in the preset period.

Specifically, the common displacement is a spatial displacement, defined as a displacement in a spatial direction, which can be denoted as D, and is three-dimensional data; in addition, radial displacement refers to displacement along the radial direction of the contour of the blood vessel wall, i.e., the distance from the contour point to the center point of the cross section, denoted as R, is one-dimensional data. The average displacement is an absolute value average value of a plurality of displacement data corresponding to each blood vessel position point in a cardiac cycle T, and the calculation formula is as follows:

where the prefix a represents an abbreviation for average, AR represents a mean value of absolute values of radial displacements, and AD represents a mean value of absolute values of spatial displacements.

The displacement space gradient is the change rate of the displacement corresponding to each vascular position point in a cardiac cycle T, and the calculation formula is as follows:

where the suffix G is an abbreviation of gradient, RG represents a change rate of radial displacement, ARG represents a mean value of the change rate of radial displacement, DG represents a change rate of spatial displacement, and ADG represents a mean value of the change rate of spatial displacement. Wherein x, y and z are displacement field data corresponding to the vascular position points respectively.

The displacement concussion index is the displacement of each vascular position point in a cardiac cycle T along with the change of time

The ratio of the amount of change to the total amount of movement path is calculated as follows:

Wherein, the prefix T is the abbreviation of time, TR represents the radial displacement oscillation index, and TD represents the space displacement oscillation index.

Optionally, determining the set of target location points in the vessel based on the spatial movement parameter includes: for any blood vessel position point, comparing one or more of average displacement, displacement space gradient and displacement concussion index of the blood vessel position point with corresponding threshold values respectively to obtain comparison results; and screening blood vessel position points meeting the motion stability condition according to the comparison result to form a target position point set.

Specifically, the vascular position points corresponding to the parameters meeting the motion stability conditions are found out through the determined spatial movement parameters of the vascular position points, and as the target position points, a set of all the target position points meeting the motion stability conditions is called a target position point set, wherein the preset conditions can be set according to actual requirements, and corresponding thresholds can be set for one or more of average displacement, displacement spatial gradient and displacement oscillation index of the vascular position points, for example, the set motion stability conditions are as follows: the average displacement of the vascular position points in the preset spatial displacement parameters is always 0.01, the displacement spatial gradient is 0.02, the threshold value corresponding to the displacement oscillation index is 0.1, if the average displacement, the displacement spatial gradient and the displacement oscillation index of a certain vascular position point are all smaller than or equal to the set threshold value, the vascular position point can be judged to be the target position point, other position points are screened according to the method until all position points are screened completely, and finally a target position point set is formed by all target position points.

According to the technical scheme, the plurality of blood vessel scanning data are respectively registered and segmented, the blood vessel position points are extracted from the processing results, the corresponding space movement parameters and the target position point set are determined through the blood vessel position point information, the problem that the movement condition of the blood vessel cannot be accurately determined is solved, the space movement condition of the position points on the blood vessel in the cardiac cycle can be more accurately determined, and the accuracy of a blood vessel image processing technology is improved.

Example two

Fig. 2 is a flowchart of a blood vessel image processing method according to a second embodiment of the present invention, where the present embodiment is an optimization scheme of the method according to the foregoing embodiment, and optionally, a plurality of blood vessel scan data are input into a pre-trained registration segmentation joint model to obtain initial segmentation data of each blood vessel scan data and a registration transformation matrix of each blood vessel scan data relative to reference blood vessel scan data, respectively; and respectively carrying out conversion processing on the corresponding initial segmentation data based on each registration transformation matrix to obtain vessel segmentation data of each vessel scanning data. As shown in fig. 2, the method includes:

s210, acquiring a plurality of blood vessel scanning data of the blood vessel to be detected in a preset period.

S220, inputting a plurality of blood vessel scanning data into a pre-trained registration segmentation joint model to obtain initial segmentation data of each blood vessel scanning data and a registration transformation matrix of each blood vessel scanning data relative to the reference blood vessel scanning data.

The initial segmentation data may be specifically understood as segmentation data obtained by processing a plurality of blood vessel scan data by a segmentation function in a registration segmentation joint model trained in advance, and may be understood as segmentation data not registered. The reference blood vessel scan data may be specifically understood as blood vessel scan data serving as registration reference, one blood vessel scan data may be randomly selected from a plurality of blood vessel scan data to serve as reference blood vessel scan data, or first-phase blood vessel scan data may be directly designated as reference blood vessel scan data, which is not limited herein, and exemplary reference blood vessel scan data may be first-phase blood vessel scan data. The registration transformation matrix is specifically understood to be a parameter for registering the initial segmentation data with the reference vessel scan data, for example, a 4×4 transformation matrix, which may be selected according to actual needs, and is not limited herein.

Specifically, the acquired multiple vessel scanning data are input into a pre-trained registration segmentation joint model, and initial segmentation data and a registration transformation matrix can be obtained through the processing of the model. And continuously adjusting and optimizing the registration transformation matrix in the training process of the registration segmentation joint model. It should be noted that, since radial deformation of the blood vessel needs to be preserved, a rigid registration method, i.e., a matrix transformation equation, is used in this embodiment.

Optionally, the training method for registering the segmentation joint model includes: acquiring a plurality of groups of blood vessel sample data, wherein any group of blood vessel sample data is a plurality of blood vessel sample data in any preset period; the following iterative training steps are executed based on the multiple sets of blood vessel sample data until the iterative stopping condition is met, so that a registration segmentation joint model with completed training is obtained:

step 1: inputting any group of blood vessel sample data into a registration segmentation joint model to be trained to obtain blood vessel segmentation data and registration results respectively corresponding to the blood vessel sample data;

step 2: and determining a joint model loss based on the label information corresponding to the blood vessel sample data, the blood vessel segmentation data corresponding to the blood vessel sample data and the registration result, and adjusting parameters of the registration segmentation joint model based on the joint model loss.

The iteration stop condition may be specifically understood as a rule condition for ending the model training set in advance, for example, the model training frequency may be set to reach a set threshold, or the model loss may be smaller than a set model loss threshold according to the model loss, which is not limited herein. The registration result can be specifically understood as a result obtained after the vessel segmentation data is rigidly registered through the registration transformation matrix. The registration transformation matrix is preset in the model in the training initial stage of the registration segmentation joint model to be trained, and is adjusted and optimized in the training process of the model. The label information is specifically understood to be attribute information corresponding to the blood vessel sample data, and may include, but is not limited to, information such as a sample name, a sample shape, a sample size, etc., and is generally labeled when the sample data is obtained, and it should be noted that not all sample data has label information.

Specifically, step 1 is executed, the acquired multiple blood vessel sample data in any preset period are used as input parameters to be input into a registration segmentation joint model to be trained, and after the model processing, the blood vessel segmentation data and the registration results corresponding to each blood vessel sample are output; and step 2 is executed, the label information corresponding to the blood vessel sample data, the blood vessel segmentation data corresponding to the blood vessel sample data and the registration result are substituted into a registration result loss function, the registration result loss, namely the joint model loss, is obtained through calculation, the joint model loss is subjected to gradient feedback, and the parameters of the registration segmentation joint model are adjusted. After the step 2 is executed, judging whether the registration segmentation joint model meets a preset iteration stop condition, if so, finishing model training, and if not, continuing to execute the step 1 until the registration segmentation joint model meeting the iteration stop condition is obtained.

Optionally, the joint model loss is determined based on the registration loss and the segmentation loss; any group of blood vessel sample data comprises reference blood vessel data and motion blood vessel data; the segmentation loss term is determined based on a first loss term formed by the blood vessel segmentation data corresponding to the reference blood vessel data and the label information corresponding to the reference blood vessel data, and/or a second loss term formed by the blood vessel segmentation data corresponding to the motion blood vessel data and the label information corresponding to the motion blood vessel data.

The first loss term refers to a loss term formed by the blood vessel segmentation data corresponding to the reference blood vessel data and the label information corresponding to the reference blood vessel data, and the second loss term refers to a loss term formed by the blood vessel segmentation data corresponding to the moving blood vessel data and the label information corresponding to the moving blood vessel data.

Specifically, if the reference blood vessel data and the motion blood vessel data are both labeled, a loss function of the segmentation result loss can be obtained by weighting and summing the first loss term and the second loss term; if only the reference vessel data is labeled, a loss function that can be lost by the first loss term as a result of the segmentation; if only the moving vessel data is labeled, a second loss term can be used as a loss function of the loss of the segmentation result; the loss function for determining the loss of the segmentation result is:

Wherein L is _S A loss function representing a loss of a determined segmentation result, I _t Representing reference vessel data, I _m Representing the data of the moving blood vessel S _t A label corresponding to the reference blood vessel data, S _m Representing the label corresponding to the moving vessel data, pred () represents the vessel segmentation result predicted by the registered segmentation joint model. It will be appreciated that pred (I _t ) Represents the vessel segmentation result corresponding to the reference vessel data, pred (I _m ) Representing the vessel segmentation result corresponding to the motion vessel data, L _S (pred(I _t ),S _t ) L is the first loss term _S (pred(I _m ),S _m ) Is the second loss term.

The registration loss is understood to be a similarity loss, which can be determined based on the degree of similarity between the registration results, and the segmentation loss can be determined based on the differences between the segmentation results. And selecting reference blood vessel data from any group of blood vessel sample data, wherein the rest blood vessel sample data are taken as motion blood vessel data. The registration loss term is based on any of the following: performing registration processing on the tag information of the moving blood vessel data based on a registration result corresponding to the moving blood vessel data to obtain registration tag information, and generating a registration loss item based on the registration tag information and the tag information of the reference blood vessel data; performing registration processing on the vessel segmentation data based on registration results corresponding to the moving vessel data to obtain registration segmentation results of the moving vessel data, and generating registration loss items based on the registration segmentation results and label information of the reference vessel data; and carrying out registration processing on the label information of the moving blood vessel data based on a registration result corresponding to the moving blood vessel data to obtain registration label information, and generating a registration loss item based on the registration label information and blood vessel segmentation data of the reference blood vessel data.

The loss function to determine registration loss is:

wherein L is _a Representing a loss function determining registration loss, I _t Representing reference vessel data, I _m Representing the data of the moving blood vessel S _t A label corresponding to the reference blood vessel data, S _m Representing the label corresponding to the moving vessel data, T representing the rigid registration transformation matrix, pred () representing the vessel segmentation result predicted by the registration segmentation joint model. It will be appreciated that S _m * T represents the registration result of the label corresponding to the moving blood vessel data. pred (I) _m ) T represents the registration result of the moving vessel data. pred (I) _t ) The blood vessel segmentation result corresponding to the reference blood vessel data is shown.

The loss function to determine the joint model loss may be:

L＝λ _s L _s +λ _a L _a

wherein L represents a loss function that determines the loss of the joint model, L _a A loss function, lambda, representing the determination of registration loss _a Weights representing a loss function determining registration loss, L _s A loss function, lambda, representing the loss of the determined segmentation result _s The weight of the loss function that determines the loss of the segmentation result is represented.

In the embodiment, registration loss and segmentation loss are determined by the label information corresponding to each blood vessel sample data, the blood vessel segmentation data corresponding to the blood vessel sample data and the registration result; the registration loss and the segmentation loss are weighted to determine joint model loss. According to the method, the registration loss and the segmentation loss are considered, the joint model loss adjustment optimization joint model is continuously determined according to the weighting processing of the registration loss and the segmentation loss in the training process of the model, and the segmentation accuracy of the registration segmentation joint model on the blood vessel image is improved.

S230, respectively converting the corresponding initial segmentation data based on each registration transformation matrix to obtain vessel segmentation data of each vessel scanning data.

Specifically, the vessel scan data is processed through the registration segmentation joint model, an initial segmentation result of each vessel scan data and a corresponding registration transformation matrix can be obtained, and the vessel segmentation data of each vessel scan data can be obtained through the transformation processing of the initial segmentation data through the registration transformation matrix.

In this embodiment, the trained registration transformation matrix is used to perform conversion processing on the initial segmentation data, so as to obtain vessel segmentation data of each vessel scan data, which is helpful for improving accuracy of the vessel segmentation data.

S240, extracting blood vessel position points based on the blood vessel segmentation data, and generating displacement field data of the blood vessel to be detected in a preset period based on the position information of the blood vessel position points in the blood vessel scanning data.

S250, determining a space movement parameter of each blood vessel position point based on the displacement field data, and determining a target position point set in the blood vessel based on the space movement parameter.

According to the technical scheme, registration loss and segmentation loss are determined based on label information corresponding to each blood vessel sample data, blood vessel segmentation data corresponding to the blood vessel sample data and registration results; and the registration loss and the segmentation loss are weighted to determine the joint model loss, and the joint model loss is continuously determined according to the weighting processing of the registration loss and the segmentation loss in the training process of the model to adjust and optimize the joint model, so that the segmentation accuracy of the registration segmentation joint model on the blood vessel image is improved.

Example III

Fig. 3 is a schematic structural diagram of a blood vessel image processing device according to a third embodiment of the present invention.

As shown in fig. 3, the apparatus includes:

the blood vessel segmentation data acquisition module 310 is configured to acquire a plurality of blood vessel scan data of a blood vessel to be detected within a preset period, and register and segment the plurality of blood vessel scan data respectively to obtain blood vessel segmentation data in each blood vessel scan data;

the displacement field data determining module 320 is configured to extract a blood vessel position point based on each blood vessel segmentation data, and generate displacement field data of a blood vessel to be detected within a preset period based on position information of the blood vessel position point in each blood vessel scan data;

the target position point set determining module 330 is configured to determine a spatial movement parameter of each blood vessel position point based on the displacement field data, and determine a target position point set in the blood vessel based on the spatial movement parameter.

Optionally, the vessel segmentation data acquisition module 310 is specifically configured to:

the preset period is a cardiac period, and the plurality of blood vessel scanning data are acquired multi-phase blood vessel scanning data obtained by carrying out blood vessel scanning based on a preset time interval in the cardiac period.

Registering and segmenting the plurality of blood vessel scanning data respectively to obtain blood vessel segmentation data in each blood vessel scanning data, wherein the registering and segmenting comprises the following steps:

The training method of the registration segmentation joint model comprises the following steps:

The joint model loss is determined based on the registration loss and the segmentation loss;

Optionally, the displacement field data determining module 320 is specifically configured to:

extracting a vessel location point based on each vessel segmentation data, comprising: for any vessel segmentation data, determining a vessel center line of the current vessel segmentation data and each center point on the vessel center line, and determining contour point cloud data of a vessel section contour in the current vessel segmentation data based on each center point, wherein the contour point cloud data is used as a vessel position point.

Determining contour point cloud data of a blood vessel cross-section contour in the current blood vessel segmentation data based on each center point comprises the following steps:

Generating displacement field data of the blood vessel to be detected in a preset period based on the position information of the blood vessel position point in each blood vessel scanning data, wherein the displacement field data comprises the following components: for any vessel position point, determining the position information of the vessel position point in each vessel segmentation data, and generating displacement field data of the vessel to be detected in a preset period based on the phase sequence of a plurality of vessel segmentation data and the position information of the vessel position point in each vessel segmentation data.

The target location point set determining module 330 is specifically configured to:

the spatial movement parameters comprise one or more of average displacement, displacement spatial gradient and displacement oscillation index; determining a spatial movement parameter for each vessel location point based on the displacement field data, including one or more of: for any vessel position point, determining average displacement based on the absolute value average value of a plurality of displacement data of the vessel position point in a preset period; determining a displacement spatial gradient based on the displacement change rate of the vascular position points in a preset period; and determining a displacement concussion index based on the ratio of the displacement variation of the vascular position point in the preset period to the total moving path in the preset period.

Determining a set of target location points in the vessel based on the spatial movement parameters, comprising: for any blood vessel position point, comparing one or more of average displacement, displacement space gradient and displacement concussion index of the blood vessel position point with corresponding threshold values respectively to obtain comparison results; and screening blood vessel position points meeting the motion stability condition according to the comparison result to form a target position point set.

The blood vessel image processing device provided by the embodiment of the invention can execute the blood vessel image processing method provided by any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.

Example IV

Fig. 4 is a schematic structural diagram of an electronic device according to a fourth embodiment of the present invention. The electronic device 10 is intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. Electronic equipment may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the inventions described and/or claimed herein.

As shown in fig. 4, the electronic device 10 includes at least one processor 11, and a memory, such as a Read Only Memory (ROM) 12, a Random Access Memory (RAM) 13, etc., communicatively connected to the at least one processor 11, in which the memory stores a computer program executable by the at least one processor, and the processor 11 may perform various appropriate actions and processes according to the computer program stored in the Read Only Memory (ROM) 12 or the computer program loaded from the storage unit 18 into the Random Access Memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 may also be stored. The processor 11, the ROM 12 and the RAM 13 are connected to each other via a bus 14. An input/output (I/O) interface 15 is also connected to bus 14.

Various components in the electronic device 10 are connected to the I/O interface 15, including: an input unit 16 such as a keyboard, a mouse, etc.; an output unit 17 such as various types of displays, speakers, and the like; a storage unit 18 such as a magnetic disk, an optical disk, or the like; and a communication unit 19 such as a network card, modem, wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information/data with other devices via a computer network, such as the internet, and/or various telecommunication networks.

The processor 11 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various processors running machine learning model algorithms, digital Signal Processors (DSPs), and any suitable processor, controller, microcontroller, etc. The processor 11 performs the respective methods and processes described above, such as a blood vessel image processing method.

In some embodiments, the vascular image processing method may be implemented as a computer program tangibly embodied on a computer-readable storage medium, such as the storage unit 18. In some embodiments, part or all of the computer program may be loaded and/or installed onto the electronic device 10 via the ROM 12 and/or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the blood vessel image processing method described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to perform the vascular image processing method in any other suitable way (e.g. by means of firmware).

Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuit systems, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), systems On Chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.

The computer program for implementing the vascular image processing method of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the computer programs, when executed by the processor, cause the functions/acts specified in the flowchart and/or block diagram block or blocks to be implemented. The computer program may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.

Example five

The fifth embodiment of the present invention also provides a computer-readable storage medium storing computer instructions for causing a processor to execute a blood vessel image processing method, the method comprising:

In the context of the present invention, a computer-readable storage medium may be a tangible medium that can contain, or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. The computer readable storage medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, the computer readable storage medium may be a machine readable signal medium. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

To provide for interaction with a user, the systems and techniques described here can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) through which a user can provide input to the electronic device. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.

The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), blockchain networks, and the internet.

The computing system may include clients and servers. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, also called a cloud computing server or a cloud host, and is a host product in a cloud computing service system, so that the defects of high management difficulty and weak service expansibility in the traditional physical hosts and VPS service are overcome.

It should be appreciated that various forms of the flows shown above may be used to reorder, add, or delete steps. For example, the steps described in the present invention may be performed in parallel, sequentially, or in a different order, so long as the desired results of the technical solution of the present invention are achieved, and the present invention is not limited herein.

The above embodiments do not limit the scope of the present invention. It will be apparent to those skilled in the art that various modifications, combinations, sub-combinations and alternatives are possible, depending on design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of the present invention.

Claims

1. A blood vessel image processing method, comprising:

acquiring a plurality of blood vessel scanning data of a blood vessel to be detected in a preset period, and respectively registering and segmenting the plurality of blood vessel scanning data to obtain blood vessel segmentation data in the blood vessel scanning data;

extracting a blood vessel position point based on each blood vessel segmentation data, and generating displacement field data of the blood vessel to be detected in the preset period based on the position information of the blood vessel position point in each blood vessel scanning data;

a spatial movement parameter of each of the vessel location points is determined based on the displacement field data, and a set of target location points in the vessel is determined based on the spatial movement parameter.

2. The method of claim 1, wherein the predetermined period is a cardiac period, and the plurality of vessel scan data is multi-phase vessel scan data obtained by performing a vessel scan based on a predetermined time interval during the cardiac period.

3. The method of claim 1, wherein registering and segmenting the plurality of vessel scan data, respectively, results in vessel segmentation data in each of the vessel scan data, comprising:

Inputting the multiple blood vessel scanning data into a pre-trained registration segmentation joint model to obtain initial segmentation data of each blood vessel scanning data and a registration transformation matrix of each blood vessel scanning data relative to reference blood vessel scanning data respectively;

4. A method according to claim 3, wherein the training method of the registration segmentation joint model comprises:

performing the following iterative training steps based on the multiple sets of blood vessel sample data until the iterative stopping condition is met to obtain a trained registration segmentation joint model:

and determining a joint model loss based on the label information corresponding to the blood vessel sample data, the blood vessel segmentation data corresponding to the blood vessel sample data and a registration result, and adjusting parameters of the registration segmentation joint model based on the joint model loss.

5. The method of claim 4, wherein the joint model loss is determined based on registration loss and segmentation loss;

the blood vessel sample data of any group comprise reference blood vessel data and motion blood vessel data;

the segmentation loss term is determined based on a first loss term formed by the blood vessel segmentation data corresponding to the reference blood vessel data and the label information corresponding to the reference blood vessel data, and/or a second loss term formed by the blood vessel segmentation data corresponding to the motion blood vessel data and the label information corresponding to the motion blood vessel data;

the registration loss term is based on any one of the following: performing registration processing on the tag information of the moving blood vessel data based on a registration result corresponding to the moving blood vessel data to obtain registration tag information, and generating a registration loss item based on the registration tag information and the tag information of the reference blood vessel data;

performing registration processing on the vessel segmentation data based on the registration result corresponding to the moving vessel data to obtain a registration segmentation result of the moving vessel data, and generating a registration loss item based on the registration segmentation result and the label information of the reference vessel data;

And carrying out registration processing on the label information of the moving blood vessel data based on a registration result corresponding to the moving blood vessel data to obtain registration label information, and generating a registration loss item based on the registration label information and the blood vessel segmentation data of the reference blood vessel data.

6. The method of claim 1, wherein said extracting a vessel location point based on each of said vessel segmentation data comprises:

for any vessel segmentation data, determining a vessel center line of the current vessel segmentation data and each center point on the vessel center line, and determining contour point cloud data of a vessel section contour in the current vessel segmentation data based on each center point, wherein the contour point cloud data is used as the vessel position point.

7. The method of claim 6, wherein said determining contour point cloud data for a vessel cross-sectional contour in the current vessel segmentation data based on each of the center points comprises:

8. The method of claim 1, wherein generating displacement field data for the vessel to be detected within the preset period based on the position information of the vessel position point in each of the vessel scan data comprises:

for any blood vessel position point, determining position information of the blood vessel position point in each blood vessel segmentation data, and generating displacement field data of the blood vessel to be detected in the preset period based on phase sequence of a plurality of blood vessel segmentation data and the position information of the blood vessel position point in each blood vessel segmentation data.

9. The method of claim 1, wherein the spatial movement parameters comprise one or more of average displacement, spatial gradient of displacement, and index of displacement concussion;

the determining a spatial movement parameter of each vessel location point based on the displacement field data includes one or more of:

for any vessel location point, determining the average displacement based on an absolute value average of a plurality of displacement data of the vessel location point within the preset period;

determining the displacement spatial gradient based on the displacement change rate of the vascular position point in the preset period;

And determining the displacement concussion index based on the ratio of the displacement variation of the vascular position point in the preset period to the total moving path in the preset period.

10. The method of claim 1, wherein the determining a set of target location points in the vessel based on the spatial movement parameter comprises:

11. A blood vessel image processing apparatus, comprising:

the blood vessel segmentation data acquisition module is used for acquiring a plurality of blood vessel scanning data of a blood vessel to be detected in a preset period, and registering and segmenting the plurality of blood vessel scanning data respectively to obtain blood vessel segmentation data in the blood vessel scanning data;

a displacement field data determining module, configured to extract a blood vessel position point based on each blood vessel segmentation data, and generate displacement field data of the blood vessel to be detected in the preset period based on position information of the blood vessel position point in each blood vessel scanning data;

12. An electronic device, the electronic device comprising:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein,,

the memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the vascular image processing method of any one of claims 1-10.

13. A computer readable storage medium, characterized in that the computer readable storage medium stores computer instructions for causing a processor to implement the blood vessel image processing method according to any one of claims 1-10 when executed.