KR20130079694A - Apparatus and method for measureing velocity vector imaging of blood vessel - Google Patents

Abstract

PURPOSE: A method and apparatus for estimating a motion in a blood vessel image are provided to find out the wall of an aorta from a multi-detector computed tomography (MDCT) image and a magnetic resonance (MR) image photographed along a cardiac cycle and to track the motion of the aorta according to time sequence. CONSTITUTION: A method for estimating a motion in a blood vessel image includes an image detecting step (100) of detecting a multi-detector computed tomography (MDCT) image or an image resonance image (MRI) with being triggered by an electrocardiogram signal; an increasing step (150) of increasing contrast with magnifying the image; an estimating step (200) of estimating the segment of an aorta; an image scaling step (250) of adjusting the image with median filtering; an aorta wall estimating step (300) of estimating the wall of the aorta in the image from the step (250); a motion tracking step (355) of detecting the motion of the wall; and a velocity mapping step (400) of mapping a velocity in the result from the step (355).

Description

Apparatus and method for measureing velocity vector imaging of blood vessel

The present invention relates to a method and apparatus for estimating motion in a blood vessel image, and more particularly, to two-dimensional multidetector computed tomography (hereinafter referred to as MDCT) and magnetic imaging of the aorta according to the cardiac activity cycle. Computer program sources for motion estimation methods, apparatus, and methods for finding aortic walls in magnetic resonance (MR) images and tracking the movement of the aorta in time order It relates to a stored recording medium.

 In order to diagnose and treat cardiovascular diseases and the like, it is necessary to measure the movement of arteries and the elasticity thereof.

 Carotid artery stiffness is measured to evaluate non-invasive local central artery stiffness. Conventional carotid artery stiffness is measured by echo tracking using ultrasound images. The motion of the distal and proximal walls of the long axis of the carotid artery is measured. Only analysis was possible.

Recently, Siemens et al. Has reported a method of measuring the motion of the carotid artery wall using speckle tracking on ultrasound images to quantify its elasticity to calculate the stiffness.

Conventional methods generally lack accuracy and precision.

In order to improve this, the present invention estimates aortic wall using ECG-gated MDCT and MR images of aorta, and estimates motion in vascular images. A method and an apparatus thereof are proposed.

In addition, the measurement method of the present invention can be applied to the same evaluation method for MDCT, MR, ultrasound vascular images.

The problem to be solved by the present invention is to find the wall of the aorta in the two-dimensional MDCT and MR images of the aorta according to the cardiac activity cycle, and to track the movement of the aorta in time sequence, motion estimation method in the vascular image To provide a recording medium storing a computer program source for, a device, and a method thereof.

Another problem to be solved by the present invention includes an image detection step triggered by an electrocardiogram signal to detect MDCT image or MRI image, aortic estimation step to estimate aortic segment, aortic wall estimation step, motion tracking step, speed mapping step The present invention provides a recording medium storing a motion estimation method, an apparatus, and a computer program source for the method in a blood vessel image.

In order to solve the technical problem of the present invention, the present invention, the image detection step of detecting the MDCT image or MRI image triggered by the ECG signal; A contrast enhancement step of enlarging the image to increase contrast; Aortic estimating step of estimating aortic segment; An image scaling step of adjusting an image including median filtering; Aortic wall estimation step of estimating the aortic wall from the image received in the image scaling step; A motion tracking step of detecting movement of the aortic wall; And a velocity mapping step of mapping a velocity from the result of the motion tracking step.

The aortic wall estimating may include marking eight initial points; Inserting interpolation at the initial points; Estimating the aortic wall boundary.

The speed mapping step may include setting a reference point; Setting a reference frame; Mapping a velocity vector; Mapping a velocity vector color.

 A recording medium storing a computer program source for the motion estimation method in the blood vessel image of the present invention.

According to the present invention, the wall of the aorta is found in the two-dimensional MDCT and MR images of the aorta according to the cardiac activity cycle, and the motion of the aorta is tracked according to the time sequence, and the motion estimation in the vascular image is highly accurate and accurate. Provided are a method, an apparatus, and a recording medium storing a computer program source for the method.

In addition, according to the present invention, the blood vessel triggered by the ECG signal, including detecting the MDCT image or MRI image, aortic estimation step for estimating aortic segment, aortic wall estimation step, motion tracking step, velocity mapping step By providing a recording medium storing a motion estimation method, an apparatus, and a computer program source for the method in the image, it is possible to obtain a more accurate and accurate result, it is easy to use.

In addition, the measurement method of the present invention can be applied to the same evaluation method for MDCT, MR, ultrasound vascular images.

1 is a schematic diagram for explaining the configuration for motion estimation in the blood vessel image according to an embodiment of the present invention.
FIG. 2 is an explanatory diagram schematically illustrating the aortic wall estimator 300 of FIG. 1.
3 is an example of the extraction of the blood vessel region in the present invention.
4 is an example of the blood vessel wall estimated after linear interpolation in the present invention.
5 is an example of the estimation and line profile of the blood vessel wall in the present invention.
6 shows examples of the initial estimation result (left) of the vascular wall and the final estimation result (right) of the vascular wall in the present invention.
7 is an example of the boundary value survey for the estimation of blood vessel wall in the present invention.
8 is an example of the amount of change of the vascular wall movement in each frame in the present invention.
9 is an example of the expression of the change in the movement of the blood vessel wall at the time of contraction (left) and relaxation (right) of the blood vessel in the present invention.
FIG. 10 is a graph illustrating the accumulation of blood vessel wall motion, an average amount of movement at each point, and a graph of comparison between the total mean and the average amount of movement of each point.
FIG. 11 is a graph showing blood vessel wall withdrawal according to the motion amount of the upper, middle and lower in the present invention.
12 is an example of the expression of the aortic vessel wall movement amount (total 20 frames) in the MDCT image in the present invention.
Figure 13 shows an example of the expression of the amount of aortic vessel wall movement in the MR image (60 frames in total).
It is an example of MRI.
15 is a flowchart of the Carotid sono wall tracking algorithm of the present invention.
16 is an example of selecting six target points in the present invention.
17 is an example of selecting 36 target points through interpolation in the present invention.
18 is an example of gray values (blue) and derivative results (red) of an image in the present invention.
19 is an example of detecting the final vessel wall motion vector in the present invention.

Hereinafter, a method and apparatus for estimating motion in a blood vessel image according to an embodiment of the present invention and a recording medium storing a computer program source for the method will be described in detail with reference to the accompanying drawings.

The present invention provides a method, apparatus, and computer program source for semi-automatically finding the walls of the aorta in two-dimensional MDCT and MR images of the aorta according to the cardiac activity cycle, and tracking the movement of the aorta over time. It relates to a recording medium storing the.

FIG. 1 is a schematic diagram illustrating a configuration for estimating motion in a blood vessel image according to an embodiment of the present invention. The image acquisition unit 100 and the enhancement of contrast 150 are shown in FIG. , Aorta segmentation unit 200, image scaling unit 250, aortic wall estimation unit 300, motion tracking unit 355, velocity mapping unit Velocity vector mapping) (400).

 The data collector 100 is triggered by an electrocardiogram signal and detects an MDCT image or an MRI or ultrasound image. In the present invention, ECG-gated MDCT images or Cine-MRA may be used to observe the movement of the aorta along the cardiac cycle.

The contrast increasing unit 150 enlarges the image to increase the contrast.

The aortic estimator 200 estimates the aortic segment. The present invention uses manual search or Hough transform to extract the aorta from the image.

The image scaling unit 250 adjusts an image including median filtering.

The aortic wall estimator 300 estimates the aortic wall from the image received by the image scaling unit 250.

The motion tracking unit 355 detects the movement of the aortic wall.

The velocity mapping unit 400 maps the velocity from the result of the motion tracking unit 355. In the present invention, the speed mapping unit 400 is configured to set a reference point, set a reference frame, map a speed vector, and map a speed vector color.

Motion estimation method in the blood vessel image, characterized in that it comprises a.

FIG. 2 is an explanatory diagram schematically illustrating the aortic wall estimator 300 of FIG. 1, wherein an initial 8 points marking 310 and an interpolation of initial of an initial point are provided. points 330, Estimation of wall boundary 350, and Sequential tracking 370.

 In the present invention, the estimation of the wall of the aorta is calculated by extracting the gradient of the image using brightness information.

The initial point marking unit 310 marks eight initial points. Herein, the present invention is not limited to the marking of the eight initial points, and the eight initial points are introduced here for convenience of description.

The interpolator 330 of the initial point inserts interpolation at the initial points.

The aortic wall boundary estimator 350 estimates the aortic wall boundary.

Tracking is sequentially performed by the tracking unit 370.

Hereinafter, the present invention will be described in brief with respect to a motion estimation method in the blood vessel image. That is, the algorithm for analyzing the motility of the blood vessel of the present invention follows the following sequence.

First, the image preprocessing process uses blood vessel images synchronized with the cardiac cycle or blood vessel images taken according to time sequence for analysis. Convert medical images in DICOM format to images in JPEG format. In addition, the image is converted into a gray scale image (see FIG. 3). The original image contrast analysis image uses a magnified (bicubic interpolation) image.

Second, the extraction process of the region of interest extracts the blood vessel region to be analyzed from the image (see FIG. 4). The blood vessel to be analyzed here is an artery. Depending on the location, the carotid artery and the ascending and descending aorta are possible.

Third, the vessel wall estimation process includes an initial estimation of the vessel wall, interpolation of the initial prediction points, and a method of analyzing the mobility of the vessels prepared for the vessel wall estimation and the vessels centered in the vessel center.

Initially, the initial estimation process of the vessel wall manually sets eight points at 45 degree intervals starting at 12 o'clock to detect the vessel wall. Mark the center of the vessel where the eight point extension lines meet inside the circle.

Next, as an interpolation process of the initial prediction points, the number of points is increased by interpolating six equally spaced points between the initial points in a linear manner. After interpolation, there are 48 points that are assumed to be the vessel wall.

Next, as a preparation for vascular wall estimation, with reference to FIG. 5, a line directed from inside to outside of the vascular wall is drawn for each of the 48 interpolated points. The length of this line can be adjusted according to the condition of the vessel wall.

Next, a process of estimating the blood vessel wall and the center of the blood vessel is described. Referring to FIG. 6, a change in the gradient of the brightness information of the pixel and the gradient of the brightness information graph and a threshold value set by the user through the blood vessel image are described. Estimate the boundaries of the vessel wall. In order to enhance the continuity in the visual representation of the vessel wall, the estimated average of the vessel wall points is moved. Finally, the point where the extension lines of the estimated points of the vessel wall are collected is used as the center point of the vessel in the frame (see FIG. 7).

The following is a process of estimating the vessel wall in the continuous image frame. Finally, the estimated set of vessel wall points is used as a reference point for the next frame vessel wall estimation. The blood vessel wall estimation algorithm is applied to the end frame of the image frame in the same manner.

Fourth, as a process of urbanization estimation of vascular motility, urbanization for motility estimation is performed after estimating all vessel walls of an image to be analyzed. At the midpoint, the frame closest to the set of vessel wall estimate points is taken as the point of contraction of the vessel. At the midpoint, the set of vessel wall estimates points the farthest frame as the point of expansion of the vessel.

The amount of movement of the blood vessel wall is based on the blood vessel wall when the blood vessel contracts, and the dots are dotted every frame until the frame at the time of expansion of the blood vessel (see FIG. 9).

The amount of movement can be expressed as the length of a line connecting a point and a point, and can also be represented by color by rank according to the length of the line segment. The method is as follows.

After plotting the degree of change of the points of the vessel wall for all frames, the average amount of movement of each point is calculated. The amount of movement is sorted by ranks from the most moving parts to the less moving points.

The point at which the amount of movement is the upper 30% is shown in red, the middle 30% is green, and for the remaining parts, the movement of the vessel wall is shown in blue (see FIG. 10).

After analysis of each frame, an image is generated as a video to observe the movement of the blood vessel wall (see FIG. 11).

12 is an example of the expression of the amount of aortic vessel wall movement in the MDCT image (20 frames in total), Figure 13 is an example of the expression of the amount of aortic vessel wall movement in the MR image (all 60 frames) in the present invention will be.

Fifth, it is possible to measure the radius and diameter of blood vessels by frame, and to visually express the change in the area of blood vessels and the quantitative changes in the amount of vascular movement in the local area. The amount of movement can be measured, and quantitative changes in vessel contraction and expansion can be observed, and when reconstructing the image by stacking on the z-axis after analyzing the cross-section of each position, three-dimensional movement of the vessel and local movement of the vessel wall can be observed. In addition, it is hoped that it will be applied to the observation of vascular wall movement acquired in most medical imaging devices.

In summary, the present invention is a development of a new image detection system using aortic MRI, in the present invention is equipped with a Sphygmocor system capable of calculating the central arterial pressure within the error range through the radial artery waveform through the transfer function (see FIG. 14). We propose an MRI imaging system that accurately and objectively reflects aortic stiffness by measuring aortic MRI and central arterial pressure simultaneously for more than 50 subjects each year.

In the present invention, the carotid artery motion evaluation using the Carotid sono wall tracking algorithm, the carotid artery stiffness is the most widely used method for evaluating the stiffness of the central artery in a non-invasive manner. As the tracking method, we use the method used in Carotid sono wall tracking algorithm.

15 is a flowchart of the Carotid sono wall tracking algorithm of the present invention.

Spirituality 4x magnification (S100), Gaussian blur (S110), initial 6 point input (S120), center point detection (S130), 30 additional points interpolation insertion (S140), image value extraction on center point-target point vector (S150), Blood vessel wall detection using the first derivative (S160), the detected blood vessel wall mean-dispersion verification (S170), the contour continuity through the median filtering (S180), the inter-frame motion vector and the target point display (S190).

The ultrasound image used for the vessel wall tracking is 312x343 in size and 8bit depth in gray scale, and one cardiac cycle is divided into 12 to 16 frames, and each frame retains the characteristics of the ultrasound image by the beamforming method of the ultrasound device. have. In particular, there are localized areas where the boundary between the vessel and the vessel wall is ambiguous, and the vessel vessel and the vessel barrier cannot be uniformly detected through the threshold. It tracks the movement of the vessel wall boundary point of the next frame based on the boundary point of the vessel wall detected in the first frame. Each image used for vascular wall tracking was enlarged because the amount of change in the vascular wall motion of the original image was too small. The image is smoothed through a Gaussian filter (radius 3.0 pixels) to eliminate the unevenness of the ultrasound image. The user specifies six points that are the boundaries of the vessel wall. This point calculates the center point of the vessel and inserts 30 additional interpolations using this center point and 6 specified points. FIG. 16 is an example of selecting six target points in the present invention, and FIG. 17 is an example of selecting 36 target points through interpolation in the present invention.

Profile the grayscale size of an image with the direction of each target point (36) from the center point as the vector direction to a point twice the distance from the center point to the target point to check the movement of blood vessels according to the contraction / relaxation of the heart Ring. In order to find the blood vessel boundary of the profiled image, the value of the large amount of change in the gray value of the image is detected through the derivative. 18 is an example of gray values (blue) and derivative results (red) of an image in the present invention.

Detecting the vessel wall using only the differential results is not significant because the local boundary of the ultrasound image and the change of gray scale value of the image are not uniform. When the gray value of the image is obtained through differentiation, the area for identifying the differential result is limited so as not to deviate greatly from the area where the actual blood vessel wall can move. Using the modified differential results, the shape of the vessel wall is verified by averaging and dispersing 36 detected target points. Singular points appearing in the unclear area are secured through median filters to obtain new target points that can provide continuity of vessel wall contours after removal. The acquired target point is used as the initial vessel wall boundary value of the next frame. The threshold value of the image is detected again using the designated initial vessel wall boundary value. It indicates the movement vector of the boundary value of the vessel wall between the previous frame and the next frame. 19 is an example of final blood vessel wall motion vector detection in the present invention.

100: data collection unit 150: contrast increasing unit
200: aortic estimator 250: image scaling unit
300: aortic wall estimation unit 310: initial point marking unit
330: Interpolation part of the initial point 350: Aortic wall boundary estimation part
355: motion tracking unit 370: sequential tracking unit
400: speed mapping unit

Claims (4)

An image detection step of detecting an MDCT image or an MRI image triggered by an ECG signal;
A contrast enhancement step of enlarging the image to increase contrast;
Aortic estimating step of estimating aortic segment;
An image scaling step of adjusting an image including median filtering;
Aortic wall estimation step of estimating the aortic wall from the image received in the image scaling step;
A motion tracking step of detecting movement of the aortic wall;
A velocity mapping step of mapping velocity from the result of the motion tracking step;
Motion estimation method in the blood vessel image, characterized in that it comprises a.
The method of claim 1,
The aortic wall estimation step
Marking eight initial points;
Inserting interpolation at the initial points;
Estimating the aortic wall boundary;
Motion estimation method in the blood vessel image, characterized in that it comprises a.
The method of claim 1,
The speed mapping step is
Setting a reference point;
Setting a reference frame;
Mapping a velocity vector;
Mapping a velocity vector color;
Motion estimation method in the blood vessel image, characterized in that it comprises a.
A recording medium storing a computer program source for the motion estimation method in any one of claims 1 to 4.

Images