WO2022120761A1 - Method and system for analyzing fluid flow in vivo, terminal, and storage medium - Google Patents

Abstract

A method and system for analyzing fluid flow in vivo, a terminal, and a storage medium. The method comprises: acquiring fluid sensitive images, at at least two different times, of a part to be examined; performing motion estimation on the fluid sensitive images by means of a motion estimation algorithm to obtain a motion field of a fluid at each position in said part, the motion field comprising a velocity vector; and calculating a nonlinear fluid velocity measurement value of at least one position in said part by means of a nonlinear velocity calculation method on the basis of the velocity vectors in the motion fields, so as to obtain a fluid flow analysis result of said part. The method allows for quick and accurate analysis of fluid flow in different parts of the whole body and in different species, and reduction of processing overhead of computers.

Description

In vivo fluid flow analysis method, system, terminal and storage medium technical field

The present application belongs to the technical field of medical image processing, and in particular relates to a method, system, terminal and storage medium for analyzing fluid flow in the body.

Background technique

To measure fluid flow, the method currently used to measure blood flow within the heart is cardiac elastography. The method measures the strain of the heart during myocardial contraction and relaxation by acquiring raw data such as tissue displacement from an echocardiogram (ultrasound of the heart). Additionally, two-dimensional slices of the heart can be imaged using standard ultrasound scans, and their Doppler flow data can be used to visualize the flow of blood in cardiac structures, for example, pulsed or continuous-wave Doppler ultrasound can be used to measure in vivo cardiac Blood flow, allowing assessment of heart valve area and function, any abnormal communication between the left and right sides of the heart, any leakage of blood through the valve (valvular regurgitation), and calculation of cardiac output and ejection fraction. However, while Doppler inspection can quantify linear flow, more detailed analysis of fluid flow exceeds the detection limit of this imaging modality.

Phase-contrast MRI is another technique that can be used to analyze blood flow, which uses the uniform motion of blood or tissue in magnetic field gradients to generate the phase-shifting properties of the MR signal. Fluid flow rates produced by phase-contrast MRI are believed to be very accurate. However, in order to achieve these speeds, the MRI machine needs to be adjusted outside its standard scanning mode, which takes a lot of extra time on top of standard MRI procedures. In addition, phase-contrast MRI is also susceptible to errors from susceptibility gradients and higher-order motions, has a small dynamic range, and requires significant computer processing overhead.

SUMMARY OF THE INVENTION

The present application provides an in vivo fluid flow analysis method, system, terminal and storage medium, aiming to solve one of the above-mentioned technical problems in the prior art at least to a certain extent.

In order to solve the above problems, the application provides the following technical solutions:

An in vivo fluid flow analysis method comprising:

Obtain at least two fluid-sensitive images of the site to be examined at different times;

Motion estimation is performed on the fluid-sensitive image by a motion estimation algorithm to obtain the motion field of each position of the fluid in the to-be-checked part; the motion field includes a velocity vector;

Based on the velocity vector in the motion field, a nonlinear velocity calculation method is used to calculate the measurement value of the nonlinear fluid velocity at at least one position in the to-be-inspected part, and the fluid flow analysis result of the to-be-inspected part is obtained.

The technical solution adopted in the embodiment of the present application further includes: the acquiring fluid-sensitive images of the to-be-inspected site at least two different times includes:

The fluid-sensitive image is segmented by active contour rendering based on the Kass-snake algorithm, and non-fluid regions in the fluid-sensitive image are excluded.

The technical solution adopted in the embodiment of the present application further includes: performing motion estimation on the fluid-sensitive image by using a motion estimation algorithm includes:

Motion estimation is performed by the pyramidal LucasKanade optical flow algorithm; the motion estimation includes:

The pixel intensity is represented by I(x, y, t), assuming that the spatiotemporal variation of the intensity signal is:

I(x,y,t)=I(x+δx,y+δy,t+δt)

In the above formula, δ represents the variable, x represents the abscissa, y represents the ordinate, and t represents the time. I(x+δx,y+δy,t+δt) represents the grayscale consistency hypothesis;

Apply chain rules to distinguish:

Where δ represents the variable, x represents the abscissa, y represents the ordinate, t represents the time, and ε represents the second-order infinitesimal term;

If the brightness of a particular point in the pattern does not change, then follow:

About the difference in t-yield:

和产量

definition

and yield

The optical flow constraint equation is:

(I _x ,I _y )·(v _x ,v _y )=-I _t

The optical flow vector has two components v _x and v _y , v _x and v _y are the velocity vectors of the optical flow at the point along the x-axis and y-axis, respectively, and the spatial gradient of the intensity is:

表示速度向量，I _t表示t时刻强度。 where ▽I represents the gradient of the intensity,

represents the velocity vector, and It represents the intensity at time _t .

The technical solution adopted in the embodiment of the present application further includes: performing motion estimation on the fluid-sensitive image by using a motion estimation algorithm includes:

Take MRI slices from axial, sagittal, and coronal scans separately and construct a 3D stacked grid or scaffold, performing three processing streams in parallel for each MRI slice from axial, sagittal, and coronal scans;

The three processing streams are initialized by a parent process that initiates the parallel processing option, and each processing stream reads MRI slices of axial, sagittal and coronal scans, respectively, and then analyzes them on a stage-by-stage basis;

Each iteration, advance to the next stage, which is initially the first stage; then, apply a motion estimation algorithm to the corresponding MRI slice to generate the intermediate motion field of the first stage;

When the final stage is reached, exiting the parallel processing, the parent process merges the intermediate motion fields of each stage, adding the intermediate vector components to form the final motion field; the motion field is a three-dimensional motion field of three-dimensional velocity vectors, including for each intersection 3D velocity vector, which is located in the middle of 3D space.

The technical solution adopted in the embodiment of the present application further includes: the calculation of the measurement value of the nonlinear fluid velocity at at least one position in the to-be-inspected part by using the nonlinear velocity calculation method based on the velocity vector in the sports field includes:

The nonlinear fluid velocity measure calculates the vorticity (ω), shear strain (Φ), and normal strain (Ψ) of the fluid according to the velocity vector in the motion field, through the vorticity (ω), shear strain (Φ) ) and the raw or average value of normal strain (Ψ) are displayed; where the vorticity (ω) represents the rotation of blood in the right atrium of the heart, the shear strain (Φ) represents the shear experienced by the blood, The normal strain (Ψ) determines the pressure experience of the blood at the local location.

The technical solution adopted in the embodiment of the present application further includes: the calculation of the measurement value of the nonlinear fluid velocity at at least one position in the to-be-inspected part by using the nonlinear velocity calculation method based on the velocity vector in the sports field includes:

Based on the velocity profile of the pixel of interest located at (i, j), the x and y components are V _x (i, j) and V _y (i, j)), respectively, N denotes the number of layers of the inner contour sampled frame, Δ _x and _Δy represents the horizontal and vertical distance between adjacent velocities, and the vorticity (ω), shear strain (Φ) and normal strain (Ψ) are calculated as:

Vorticity (ω):

Shear strain (Φ):

Normal strain (Ψ):

The technical solution adopted in the embodiment of the present application further includes: after calculating the measured value of the nonlinear fluid velocity at at least one position in the to-be-inspected part by using the nonlinear velocity calculation method, the method further includes:

A representation or motion field of the nonlinear fluid velocity measure is overlaid on the fluid-sensitive image and displayed.

Another technical solution adopted in the embodiment of the present application is: an in-vivo fluid flow analysis system, comprising:

Magnetic resonance imager: used to obtain at least two MR images of the site to be examined at different times;

Motion estimation element: used to perform motion estimation on the fluid-sensitive image through a motion estimation algorithm to obtain the motion field of the fluid at each position in the to-be-inspected part;

Calculation element: used to calculate the measurement value of the nonlinear fluid velocity at at least one position in the to-be-inspected part based on the velocity vector in the motion field by using a nonlinear velocity calculation method to obtain the fluid flow analysis result of the to-be-inspected part ;

Display element: for superimposing and displaying a representation or motion field of the nonlinear fluid velocity measure on the fluid-sensitive image.

Another technical solution adopted by the embodiments of the present application is: a terminal, the terminal includes a processor and a memory coupled to the processor, wherein,

the memory stores program instructions for implementing the in vivo fluid flow analysis method;

The processor is configured to execute the program instructions stored in the memory to control in vivo fluid flow analysis.

Another technical solution adopted by the embodiments of the present application is: a storage medium storing program instructions executable by a processor, where the program instructions are used to execute the in-vivo fluid flow analysis method.

Compared with the prior art, the beneficial effects of the embodiments of the present application are: the method, system, terminal and storage medium for analyzing fluid flow in the body of the embodiments of the present application obtain fluid-sensitive images of the body by using various mechanisms, and make the motion estimation algorithm The image generates a motion field from which a measure of non-linear fluid flow at at least one location within the body is calculated and displayed by superimposing a representation of the calculated measure on the body image. The embodiments of the present application can quickly and accurately analyze the fluid flow of different parts and species in the whole body, and reduce the computer processing overhead.

Description of drawings

Fig. 1 is the flow chart of the in vivo fluid flow analysis method of the embodiment of the present application;

2 is a schematic diagram of an MR image according to an embodiment of the present application;

3 is a schematic diagram of the operation of the pyramidal LucasKanade optical flow algorithm performed by two pixel groups at different times in the embodiment of the present application;

FIG. 4 is a display mode of the sports field according to the embodiment of the present application, wherein the high turbulence area is displayed as a darker area on the image;

5 is a schematic view of a slice of a group of MR images in an orthogonal plane in a three-dimensional space according to an embodiment of the present application;

6 is a schematic diagram of a three-dimensional sports field generation process according to an embodiment of the present application;

Fig. 7 is the histogram of the nonlinear fluid velocity measurement value of the embodiment of the present application;

8 is a graph of the average value of nonlinear fluid velocity measurement values calculated at different times in an embodiment of the present application;

9 is a schematic diagram of superimposed display of the motion field generated by the pyramidal LucasKanade optical flow algorithm and the MR image shown in FIG. 2 according to an embodiment of the present application;

10 is a schematic structural diagram of an in vivo fluid flow analysis system according to an embodiment of the present application;

FIG. 11 is a schematic structural diagram of a terminal according to an embodiment of the present application;

FIG. 12 is a schematic structural diagram of a storage medium according to an embodiment of the present application.

Detailed ways

In order to make the purpose, technical solutions and advantages of the present application more clearly understood, the present application will be described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, but not to limit the present application.

Please refer to FIG. 1 , which is a flowchart of an in vivo fluid flow analysis method according to an embodiment of the present application. The in vivo fluid flow analysis method of the embodiment of the present application includes the following steps:

S10: Acquiring at least two fluid-sensitive images of the site to be inspected at different times;

In this step, the fluid-sensitive images include different types of images of different parts of the body, such as the heart. For the convenience of description, the embodiment of the present application only takes an MR image obtained from a CMR (cardiac magnetic resonance examination) image scan as an example. Images were acquired by performing steady-state free-precession cine-MR imaging using consecutive slices in a short-axis view across the heart, with retrospective gating to acquire 25 stages per slice (from T=0 to 24) , to provide 25 MR images of the heart at different times. Specifically, as shown in FIG. 2 , which is a schematic diagram of an MR image according to an embodiment of the present application, which shows the outline of a chamber (right atrium) of a human heart in four different stages of a cardiac cycle.

It can be understood that the fluid-sensitive image in this embodiment of the present application can be stored in any image format such as a bitmap format, and in actual operation, a pre-stored image can also be directly read for fluid analysis without real-time acquisition. In addition, the embodiments of the present application can be applied to different types of MR images such as phase contrast, gradient echo, or markers, using intensity contrast on standard MR images without using markers for MRI procedures. In addition, the embodiments of the present application can also be applied to other types of fluid-sensitive images such as particle image velocimetry (PIV) images.

S20: segment the fluid-sensitive image to exclude non-fluid regions in the fluid-sensitive image;

In this step, in order to avoid the influence of other tissues on the fluid analysis, in this embodiment of the present application, the active contour drawing based on the Kass-snake algorithm is performed, and a two-dimensional contour is placed to segment the heart wall, and the two-dimensional contour forms a computational elasticity in the cardiac cavity. Wall, migration of contour nodes from their origin to the wall region is performed based on an energy minimization algorithm to define the inner boundary of the heart in order to exclude non-regions such as walls and substances of the body.

In the embodiment of the present application, the fluid-sensitive image segmentation method is specifically:

The contour of the heart chamber is described as an energy function _Econtour , receiving information from the previous contour, and applying an energy balance based on the contour's internal energy E _int and external energy E _ext , respectively, to redefine the contour representation. The fitted contour is the one corresponding to the minimum value of this energy:

E _contour =∫E _int +∫E _ext (1)

The initial curve can be anywhere in the MR image, with automatic detection of internal contours and manual tracking of the inner walls of the heart chambers if poor segmentation is caused by over-expansion of the elastic contours. Due to the semi-automatic nature of this segmentation, contour tracking can be preprocessed.

S30: Perform motion estimation on the segmented fluid-sensitive image through a motion estimation algorithm to obtain motion fields of fluid at various positions in the heart;

In this step, the motion estimation algorithm used in this embodiment of the present application is a pyramid-shaped LucasKanade optical flow algorithm. Based on the differences of MR images at different times, top-down flow estimation is performed using an image pyramid, where the vertices represent the CMR images at a coarse scale (useful for obtaining a global representation of fluid flow), and the computation results from this level are Pass on to the next, and the process proceeds based on the estimated flow at the previous scale until a substantially fine (eg, single pixel) resolution scale is reached to estimate the motion field of the fluid at various locations within the body.

MR images are fluid-sensitive, and the intensity contrast of turbulent and laminar blood flow is characteristic of MRI scans, allowing the visualization of blood motion based on the intensity of movement in the image due to signal voids caused by dephasing of nuclear spins. The regions of high turbulence in the MRI scan are darker than the regions of low turbulence, as shown in Figure 3, a schematic diagram of the operation of the pyramidal LucasKanade optical flow algorithm for two pixel groups at different times. In (a), a very simple image is divided into four quadrants or regions, where the lower left region is darker than the others. In (b), taken later, the upper right area is darker, and then the object producing the dark area (i.e. the turbulent area) is moved diagonally across the image. The motion field includes velocity vectors in the turbulent region of the fluid. As shown in Figure 4, it is a display mode of the sports field, in which the high turbulence area is displayed as a darker area on the image, and the intensity of this area corresponds to the magnitude of the velocity vector. Since a motion field is a collection of velocity vectors at different points throughout the body, typically, each of these points corresponds to a pixel of at least one image, embodiments of the present application only track the movement of the turbulent region, ie the motion field includes only the movement of the turbulent region. Velocity vector without including every pixel velocity vector to save computational load.

The pyramidal LucasKanade optical flow algorithm can be applied to regions of various scales, ranging from fine pixel resolution (i.e. a single pixel) to very coarse pixel resolution (i.e. large groups of regions). Taking fine pixel resolution as an example, the implementation process of the pyramidal LucasKanade optical flow algorithm includes:

The pixel intensity is represented by I(x, y, t), assuming that the spatiotemporal variation of the intensity signal is:

I(x,y,t)=I(x+δx,y+δy,t+δt) (2)

In formula (2), δ represents the variable, x represents the abscissa, y represents the ordinate, and t represents the time. I(x+δx, y+δy, t+δt) represents the grayscale consistency hypothesis.

Expand the Taylor formula on the right side of Equation (2), namely:

In formula (3), δ represents the variable, x represents the abscissa, y represents the ordinate, t represents the time, and ε represents the second-order infinitesimal term. If the brightness of a particular point in the pattern does not change, then follow:

About the difference in t-yield:

和产量

definition

and yield

Therefore, the optical flow constraint equation can be rewritten as:

(I _x ,I _y )·(v _x ,v _y )=-I _t (6)

In formula (6), v _x and v _y are the velocity vectors of the optical flow at the point along the x-axis and y-axis directions, respectively.

The optical flow vector has two components v _x and v _y , which are used to describe the motion of feature points in the x and y directions, and the spatial gradient of the intensity is expressed by:

表示速度向量，I _t表示t时刻强度。 In formula (7), ▽I represents the gradient of intensity,

represents the velocity vector, and It represents the intensity at time _t .

Therefore, a linearized version of the luminance constancy assumption yields optical flow constraints.

It can be understood that in other embodiments of the present application, Horn-SchunckOF algorithm (a global method that introduces global constraints on smoothness) or other motion estimation algorithms of the block matching method may also be used.

Fluid flow is generally not limited to a single plane. Therefore, to more accurately analyze fluid flow within the heart, when executing motion estimation algorithms, MRI slices from axial, sagittal, and coronal scans, respectively, are taken and a 3D stacked grid or scaffold is constructed, and a 3D motion field is generated. Specifically, as shown in FIG. 5 , it is a schematic diagram of slices of a group of MR images in an orthogonal plane in a three-dimensional space. Multiple planes intersect at different locations in the body, the intercept points of the three image slices are represented by spherical anchors, and the axial planes are hidden to reveal spherical anchors.

Please refer to FIG. 6 , which is a schematic diagram of a three-dimensional sports field generation process according to an embodiment of the present application, which shows a three-dimensional sports field including X, Y and Z velocity vectors from interception points. The three-dimensional velocity vector includes the sum of orthogonal velocity component vectors in three-dimensional space through the intersection of the slices. That is: for each intercept point, compute a composite velocity vector based on the addition of orthogonal velocity components from the two-dimensional slice. Three processing streams are executed in parallel for each MRI slice of the axial, sagittal and coronal scans. The three processing streams are initialized by the parent process, which initiates the parallel processing option. Each processing stream reads MRI slices from axial, sagittal, and coronal scans, respectively, and then analyzes them on a phase-by-phase basis for each phase of the cardiac cycle. Each iteration, advance to the next stage, which will initially be the first stage. Then, a motion estimation algorithm is applied to the corresponding MRI slices to generate the intermediate motion fields of the first stage. When the final stage has been analyzed, the parallel processing is exited, and the parent process then merges the intermediate motion fields of each stage, adding the intermediate vector components to form the final motion field (for each phase). The motion field is a three-dimensional motion field of three-dimensional velocity vectors, including a three-dimensional velocity vector for each intersection, which is located in the middle of three-dimensional space.

Based on the above, flow visualization using MRI scans is fast, non-invasive, and unlimited due to the opacity and motion of the body, and (for medical applications) has the advantage of using commonly used techniques. It will be appreciated that the present invention is applicable to any imaging technique. The intensity contrast of turbulent and laminar blood flow is a feature of certain types of MRI scans that allow visualization of blood motion based on the intensity of movement in the image due to signal voids caused by dephasing of nuclear spins.

S40: Based on the velocity vector in the motion field, a nonlinear velocity calculation method is used to calculate the measurement value of the nonlinear fluid velocity at at least one position in the to-be-inspected part to obtain the fluid flow analysis result of the to-be-inspected part;

In this step, the nonlinear fluid velocity measurement value calculates the nonlinear flow such as vorticity (ω), shear strain (Φ) and normal strain (Ψ) of the fluid according to the velocity vector in the motion field. Raw or average values of shear strain (Φ) and normal strain (Ψ) are displayed; where vorticity (ω) represents the rotation of blood in the right atrium of the heart, shear strain (Φ) represents the shear experienced by the blood, The normal strain (Ψ) determines the pressure experience of the blood at the local location. Velocity contour based on the pixel of interest located at (i, j) (its x and y components are V _x (i, j) and V _y (i, j), respectively), n denotes the number of layers of the contour sampled within the frame, Δ _x and _Δy represent the horizontal and vertical distances between adjacent velocities, and the vorticity (ω), shear strain (Φ) and normal strain (Ψ) are calculated as follows:

Vorticity (ω):

Shear strain (Φ):

Normal strain (Ψ):

Vorticity is a component of turbulence, but turbulence also includes random or chaotic fluid flow. In the embodiment of the present application, the statistical quantification of the nonlinear fluid velocity measurement value may also be performed by calculating other nonlinear flow rates such as the number, size or direction of eddy currents or turbulent flow regions in the body. As shown in FIG. 7 , it is a histogram of the nonlinear fluid velocity measurement value of the embodiment of the present application. This vorticity histogram can give guidance on the general propagation of vortices in the body, given that the vorticity has been calculated for many locations in the body. Figure 8 is a graph of the mean values of nonlinear fluid velocity measures calculated at different times, which plots a basic line graph of mean calculated vorticity, shear strain, and normal strain within the heart at different stages of the heart. In clinical applications, the shape of the figure can be compared to that of a healthy heart to look for abnormalities.

S50: Superimpose the representation or motion field of the nonlinear fluid velocity measurement value on the fluid-sensitive image and display it;

In this step, the nonlinear fluid velocity measure can be displayed by overlaying a representation of the value at the corresponding location in the image, which can include "dots" of specific intensity or specific color (depending on the value of the measure). In addition, the motion field can be superimposed on the image simultaneously or individually, with each point of the motion field displayed at a corresponding location on the image. As shown in FIG. 9 , it is a schematic diagram showing the superposition of the motion field generated by the pyramidal LucasKanade optical flow algorithm and the MR image (T=8) shown in FIG. 2 . In Figure 9, an outline is shown on the image, which shows the flow pattern of the blood in the right atrium at the current stage, and shows a region of similar magnitude of vorticity. The arrows in the figure represent the velocity vectors in the sports field, and the length of the arrows corresponds to the magnitude of the velocity vectors. Contours can also be used to show areas with similar shear or normal strain, allowing easy visualization of the center of the main vortex.

It can be understood that there are various forms of information display, such as the use of color scales, etc. The color of the area superimposed on the image can indicate the magnitude and direction of vorticity, and the color can distinguish clockwise rotation (for example, red) and anti-blood. Clockwise movement (eg blue).

Based on the above, the in-vivo fluid flow analysis method of the embodiment of the present application obtains a fluid-sensitive image of the body by using various mechanisms, generates a motion field from the image through a motion estimation algorithm, and calculates a measurement value of nonlinear fluid flow at at least one position in the body according to the motion field. , by superimposing a representation of the calculated metric value onto the body image and displaying it. It can be used for auxiliary diagnosis of various diseases related to the pathogenesis of cardiovascular disease, such as atherosclerosis, arterial disease, heart defect or turbulent blood flow. For example, in a septal defect (a ventricle or atrium), oxygenated blood is forced from the left side of the heart to the right side of the heart through a hole in the septum, causing excess blood to enter the lungs (via the pulmonary artery) and drain the body tissue (via the aorta). Too little oxygen is not enough to distribute to other parts of the body, which can lead to abnormal blood flow patterns in the heart, and the analysis results of this application can allow clinicians to understand the abnormal blood flow patterns, so as to quickly take effective measures to patients for rescue.

It is understood that the embodiments of the present application can be used for fluid (eg, cerebrospinal fluid) flow analysis in different parts and species throughout the body. Furthermore, the present application has obvious applicability to the design and testing of biomedical devices such as artificial hearts or mechanical heart valves, can be used for the design and optimization of artificial valves, and can also be used to identify risks that may arise after heart valve transplantation. It can also be widely used in non-biological applications that require flow analysis, such as the analysis of fluid flow in manufacturing through or in machinery from an engineering perspective, airflow analysis in aeronautics (e.g. reducing airflow turbulence on aircraft structures), piping or Fluid flow in pipes (e.g. optimizing ink flow efficiency in printers), etc. In fact, the present invention can be applied to any area (PIV, Particle Image Velocimetry, Particle Image Velocimetry) currently used.

The embodiments of the present application can be applied to cardiac analysis after cardiac surgery, to determine the success of the patient's surgery and to help management decide to stabilize the heart disease. In the event of heart valve failure the heart may require more energy to pump blood, vortices are energy retaining structures and are observed in normal heart chambers, increasing the efficiency of the heart. The flow information produced by this application can be used to examine the amount of energy wasted by the heart, which may need to pump blood through abnormal heart valves to maintain the desired circulation in the body. The present invention provides the potential for non-invasive flow visualization and quantification in cardiac structures, such as in vivo natural and bioprosthetic heart valves changing their spatial position over time in a beating heart.

Please refer to FIG. 10 , which is a schematic structural diagram of an in vivo fluid flow analysis system according to an embodiment of the present application. First, MR images taken at two or more different times are acquired by the magnetic resonance imager 210 or from the scanner 220 or disk drive 230, and after being received by the receiver 240, resides on the processor. In the event that the image requires segmentation, non-fluid structures are excluded from the MR image by segmentation element 250 . Segmentation element 250 may allow a user to assist in the segmentation process via an input device such as keyboard 260 or mouse 270 . The motion estimation element 320 then applies a motion estimation algorithm to the MR images to generate a motion field.

The sports field may be displayed along with the MR images by display element 280, which may communicate with projector 290 or monitor 300 to display the sports field and/or images in a set manner.

The system may also include a computing element 310 for computing a measure of non-linear fluid flow in at least one location within the body from the sports field, which measure may also be displayed by the display element 280 . Display element 280 may be adapted to display motion fields and measures of nonlinear fluid flow in three dimensions.

Experiments and applications have shown that the present application can be used for the specific application of analyzing blood flow within the human heart, but also for analyzing fluid flow throughout the rest of the body or in different species, and for biomedical devices such as artificial hearts or mechanical heart valves The design and testing have obvious applicability.

Please refer to FIG. 11 , which is a schematic structural diagram of a terminal according to an embodiment of the present application. The terminal 50 includes a processor 51 and a memory 52 coupled to the processor 51 .

The memory 52 stores program instructions for implementing the in vivo fluid flow analysis method described above.

The processor 51 is configured to execute program instructions stored in the memory 52 to control the in vivo fluid flow analysis.

The processor 51 may also be referred to as a CPU (Central Processing Unit, central processing unit). The processor 51 may be an integrated circuit chip with signal processing capability. The processor 51 may also be a general purpose processor, digital signal processor (DSP), application specific integrated circuit (ASIC), off-the-shelf programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware component . A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

Please refer to FIG. 12 , which is a schematic structural diagram of a storage medium according to an embodiment of the present application. The storage medium of this embodiment of the present application stores a program file 61 capable of implementing all the above methods, wherein the program file 61 may be stored in the above-mentioned storage medium in the form of a software product, and includes several instructions to enable a computer device (which may It is a personal computer, a server, or a network device, etc.) or a processor that executes all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, Read-Only Memory (ROM, Read-Only Memory), Random Access Memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program codes , or terminal devices such as computers, servers, mobile phones, and tablets.

The above description of the disclosed embodiments enables any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined in this application may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (10)

A method for analyzing fluid flow in vivo, comprising: Obtain at least two fluid-sensitive images of the site to be examined at different times; Motion estimation is performed on the fluid-sensitive image by a motion estimation algorithm to obtain the motion field of the fluid at each position in the to-be-inspected part; the motion field includes a velocity vector; Based on the velocity vector in the motion field, a nonlinear velocity calculation method is used to calculate the measurement value of the nonlinear fluid velocity at at least one position in the to-be-inspected part, and the fluid flow analysis result of the to-be-inspected part is obtained. The in-vivo fluid flow analysis method according to claim 1, wherein the acquiring fluid-sensitive images of the to-be-examined site at least two different times comprises: The fluid-sensitive image is segmented by active contour rendering based on the Kass-snake algorithm, and non-fluid regions in the fluid-sensitive image are excluded. The in-vivo fluid flow analysis method according to claim 1, wherein the performing motion estimation on the fluid-sensitive image by a motion estimation algorithm comprises: Motion estimation is performed by the pyramidal LucasKanade optical flow algorithm; the motion estimation includes: The pixel intensity is represented by I(x, y, t), assuming that the spatiotemporal variation of the intensity signal is: I(x,y,t)=I(x+δx,y+δy,t+δt) In the above formula, δ represents the variable, x represents the abscissa, y represents the ordinate, t represents the time, and I(x+δx, y+δy, t+δt) represents the grayscale consistency hypothesis; Apply chain rules to distinguish:

Where δ represents the variable, x represents the abscissa, y represents the ordinate, t represents the time, and ε represents the second-order infinitesimal term; If the brightness of a particular point in the pattern does not change, then follow:

About the difference in t-yield:

definition

and yield

The optical flow constraint equation is: (I _x ,I _y )·(v _x ,v _y )=-I _t The optical flow vector has two components v _x and v _y , v _x and v _y are the velocity vectors of the optical flow at the point along the x-axis and y-axis, respectively, and the spatial gradient of the intensity is:

in,

represents the gradient of intensity,

represents the velocity vector, and It represents the intensity at time _t . The in-vivo fluid flow analysis method according to claim 3, wherein the performing motion estimation on the fluid-sensitive image by a motion estimation algorithm comprises: Take MRI slices from axial, sagittal, and coronal scans separately and construct a 3D stacked grid or scaffold, performing three processing streams in parallel for each MRI slice from axial, sagittal, and coronal scans; The three processing streams are initialized by a parent process that initiates the parallel processing option, and each processing stream reads MRI slices of axial, sagittal and coronal scans, respectively, and then analyzes them on a stage-by-stage basis; Each iteration, advance to the next stage, which is initially the first stage; then, apply a motion estimation algorithm to the corresponding MRI slice to generate the intermediate motion field of the first stage; When the final stage is reached, exiting the parallel processing, the parent process merges the intermediate motion fields of each stage, adding the intermediate vector components to form the final motion field; the motion field is a three-dimensional motion field of three-dimensional velocity vectors, including for each intersection 3D velocity vector, which is located in the middle of 3D space. The method for analyzing fluid flow in the body according to claim 4, wherein the non-linear velocity calculation method based on the velocity vector in the motion field is used to calculate the measurement value of the nonlinear fluid velocity at at least one position in the site to be inspected, comprising: : The nonlinear fluid velocity measure calculates the vorticity (ω), shear strain (Φ) and normal strain (Ψ) of the fluid according to the velocity vector in the motion field, by vorticity (ω), shear strain (Φ) ) and the raw or average value of normal strain (Ψ) are displayed; where the vorticity (ω) represents the rotation of blood in the right atrium of the heart, the shear strain (Φ) represents the shear experienced by the blood, The normal strain (Ψ) determines the pressure experience of the blood at the local location. The method for analyzing fluid flow in vivo according to claim 5, characterized in that, based on the velocity vector in the motion field, using a nonlinear velocity calculation method to calculate the measurement value of the nonlinear fluid velocity at at least one position in the site to be inspected comprises the following steps: : Based on the velocity profile of the pixel of interest located at (i, j), the x and y components are V _x (i, j) and V _y (i, j)), respectively, N denotes the number of layers of the inner contour sampled frame, Δ _x and _Δy represents the horizontal and vertical distance between adjacent velocities, and the vorticity (ω), shear strain (Φ) and normal strain (Ψ) are calculated as: Vorticity (ω):

Shear strain (Φ):

Normal strain (Ψ):

The method for analyzing fluid flow in the body according to any one of claims 1 to 6, characterized in that, after calculating the measurement value of the nonlinear fluid velocity at at least one position in the to-be-examined site by using a nonlinear velocity calculation method, further include: A representation or motion field of the nonlinear fluid velocity measure is overlaid on the fluid-sensitive image and displayed. An in vivo fluid flow analysis system, characterized in that it includes: Magnetic resonance imager: used to obtain at least two MR images of the site to be examined at different times; Motion estimation element: used to perform motion estimation on the fluid-sensitive image through a motion estimation algorithm to obtain the motion field of the fluid at each position in the to-be-inspected part; Calculation element: used to calculate the measurement value of the nonlinear fluid velocity at at least one position in the to-be-inspected part based on the velocity vector in the motion field by using a nonlinear velocity calculation method to obtain the fluid flow analysis result of the to-be-inspected part ; Display element: for superimposing and displaying a representation or motion field of the nonlinear fluid velocity measure on the fluid-sensitive image. A terminal, characterized in that the terminal includes a processor and a memory coupled to the processor, wherein, The memory stores program instructions for implementing the in vivo fluid flow analysis method according to any one of claims 1-7; The processor is configured to execute the program instructions stored in the memory to control in vivo fluid flow analysis. A storage medium, characterized in that it stores program instructions executable by a processor, and the program instructions are used to execute the method for analyzing fluid flow in the body according to any one of claims 1 to 7.

Images