WO2016068395A1 - Method for detecting blood vessels on basis of medical image, and apparatus thereof - Google Patents

Method for detecting blood vessels on basis of medical image, and apparatus thereof Download PDF

Info

Publication number
WO2016068395A1
WO2016068395A1 PCT/KR2015/000946 KR2015000946W WO2016068395A1 WO 2016068395 A1 WO2016068395 A1 WO 2016068395A1 KR 2015000946 W KR2015000946 W KR 2015000946W WO 2016068395 A1 WO2016068395 A1 WO 2016068395A1
Authority
WO
WIPO (PCT)
Prior art keywords
blood vessel
direction
set
radial
brightness value
Prior art date
Application number
PCT/KR2015/000946
Other languages
French (fr)
Korean (ko)
Inventor
장혁재
심학준
한동진
전병환
장영걸
홍영택
정성희
하성민
Original Assignee
연세대학교 산학협력단
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority to KR1020140148520A priority Critical patent/KR101630231B1/en
Priority to KR10-2014-0148520 priority
Priority to KR1020140148531A priority patent/KR101579902B1/en
Priority to KR10-2014-0148531 priority
Application filed by 연세대학교 산학협력단 filed Critical 연세대학교 산학협력단
Publication of WO2016068395A1 publication Critical patent/WO2016068395A1/en

Links

Images

Classifications

    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Detecting, measuring or recording for diagnostic purposes; Identification of persons
    • A61B5/02Detecting, measuring or recording pulse, heart rate, blood pressure or blood flow; Combined pulse/heart-rate/blood pressure determination; Evaluating a cardiovascular condition not otherwise provided for, e.g. using combinations of techniques provided for in this group with electrocardiography or electroauscultation; Heart catheters for measuring blood pressure
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
    • A61B6/02Devices for diagnosis sequentially in different planes; Stereoscopic radiation diagnosis
    • A61B6/03Computerised tomographs

Abstract

The present invention relates to a method for detecting blood vessels on the basis of a medical image, and an apparatus thereof, the method comprising: a first step of generating a parameter space including blood vessels seen on a medical image obtained through angiography; a second step of generating a plurality of curves by using a bezier curve on the basis of a normal vector placed in the blood vessel direction, while variably controlling the size of the generated parameter space; and a third step of selecting at least one curve among the plurality of generated curves.

Description

Medical imaging-based vessel detection method and apparatus

The present invention relates to a medical image based on blood vessel detection method, and relates to an apparatus, and more particularly medical image based on blood vessel detection method for detecting a blood vessel on the basis of the medical image obtained through angiography and apparatus. The present invention claims the benefit of the filing date favors the Korea Patent Application No. 10-2014-0148531 filed on Oct. 29 as the same as the Korea Patent Application No. 10-2014-0148520, filed in 2014, and all its contents herein It is included.

The prior cardiovascular tracking method uses a method to determine how to estimate, using the blood vessel cross-section of the circular nature of the blood vessels in medical imaging that is, a circle or sphere, the presence of the blood vessel based on the model, such as an oval or tawongu mainly It was.

This method, however, was in most cases dependent on the quality of medical images obtained in a controlled environment. In addition, it becomes narrower toward down to the lower end of the blood vessel cross-section is looking at is thought to be cylindrical since the convergence in the form of dots than a circle, and continues to track there is a problem actually occurs one who is not found even though there is a blood vessel.

In this regard, Korea Patent Publication No. 2012-0119523 discloses a is an "apparatus for blood vessel detection position" disclosed.

Korea Patent Publication No. 2011-0077740 discloses a method for tracking the three-dimensional blood vessel CTA (Computed Tomography Angiography) images. How to track these vessels it was very difficult to figure out exactly how to stop at some point stopped at the foot vessels, because the non-linear predictive data.

Accordingly, in recent years it has used how to terminate it became similar to the brightness value and the background value of the blood vessel. However, this method has a problem that is not easy to sensitively influenced by the value of the extracted background exit in the proper position blossomed false positives (False positive), depending on other areas, such as ventricular myocardium and the surrounding blood vessels.

The invention medical images to generate a plurality of curves to as an invention to solve the above problems, based on the, normal vector positioned in blood vessel direction to detect a blood vessel present in the medical image obtained through angiography to provide a vessel based on the detection method and apparatus it is an object.

Further, the invention provides a selected blood vessel of the generated plurality of curves and the most likely one of the medical image based on blood vessel blood vessel detection method and apparatus it is an object.

In addition, as one of this invention it is any of the vessels direction in which the vessel wall or general vascular direction or plaque (plaque) exists based on the difference of the luminance intensity values ​​of the values ​​and the spherical coordinate system normal to a radial line of the vector direction to take advantage of the extraction from the origin of a blood vessel to provide a medical image is determined based on blood vessel detection method and apparatus for it is an object.

The present invention also provides a case that it is determined in blood vessel direction under a normal blood vessel direction or plaque is present, group if the number of elements belonging to a predetermined set of conditions converge to zero medical image based on blood vessel detection method to end a blood vessel is detected and an apparatus to have its purpose.

Medical images based blood vessel detection method according to the present invention for achieving the above object is a first step of creating a parameter space (parameter space) including a blood vessel shown in the medical image obtained through angiography; And variably adjusting the size of the generated parameter space, the second step using a Bezier curve (bezier curve) relative to the normal vector direction is placed in blood vessel for generating a plurality of curves; And a third step of selecting at least one curve of the generated plurality of curved lines; characterized in that it comprises a.

Further, a first step of creating a parameter space containing the blood vessel shown in the medical image obtained through angiography is characterized in that to generate the cylindrical space of the parameters, including blood vessels.

Further, while variably adjusting the size of the generated parameter space, in the second step Bezier relative to the normal line vector placed the vessel direction by using a curved (bezier curve) to generate a plurality of curve, radius of the parameter space a characterized in that the variably controlled.

Further, while variably adjusting the size of the generated parameter space, Bezier relative to the normal line vector placed the vessel direction by using a curved (bezier curve) to generate a plurality of curves, the start and end points of the normal vectors a second step of fixing; By using the Bezier curve calculating repeatedly a first point and a second point from the surface of the parameter space; And by the product of the set of the calculated first point of the second set of points to produce a plurality of curved lines; it characterized in that it comprises a.

Further, a third step of selecting at least one curve of the generated plurality of curves is characterized by selecting at least one of the curves is determined that one of the plurality of curves similar to the blood vessel.

Medical images based blood vessel detection method according to the above-described present invention, by based on the difference of the luminance intensity values ​​of the values ​​and the spherical coordinate system normal to a radial line of the vector direction to take advantage of the extraction from the origin of the vascular vessel wall, or general vascular direction or plaque ( plaque) is present and further comprising a blood detection end time determination step of determining any one of a vascular direction, exit the blood vessel detection time determination process, brightness specify the origin of a blood vessel and, inside the average brightness value of the blood vessel and the outer average a first step for extracting a value; A second step of extracting a plurality of normal vectors facing the direction of the coordinate values ​​and brightness values ​​of the radial line relative to the origin of the blood vessel; A third step of determining the direction of the blood vessel based on the difference of the inner average brightness value and the brightness value of said radial line; A fourth step of the operation if it is in the set of predetermined conditions with respect to the radial lines, including a radial straight line corresponding to the direction in which the blood vessel general direction or the plaque exists on the set group set; And a fifth step of detecting when to exit the vessel but satisfies the condition of the group set is set, the number of elements belonging to the group set set converge to zero; may include.

Further, in the second step of extracting a plurality of normal vectors facing the direction of the coordinate values ​​and brightness values ​​of the radial line relative to the origin of the blood vessel, the coordinate values ​​of the radial line it is characterized by using a spherical coordinate system.

Further, when in the third step of determining the direction of the blood vessel based on the difference of the inner average brightness value, brightness values ​​of the radial line, the difference between the brightness values ​​of the internal average brightness value and the radial line of the positive (+) vascular direction to the case of, the said radial straight line common vascular direction or plaque present -, wherein a radial line is determined to be the blood vessel wall, and the difference between the brightness values ​​of the internal average brightness value and the radial straight line zero or negative () It characterized in that it is determined to be.

Further, in the fourth step to include a radial straight line by operation if it is in the set of conditions predetermined with respect to the radial straight lines, corresponding to the vessels direction of the normal vascular direction or plaque exists based on predetermined set, set the preset the condition or if the absolute value of the difference between the brightness value of the brightness value against the inner average brightness value and the radial line of the radial straight line is smaller than the absolute value of the difference between the inner average gray level and outside the average brightness value of the blood vessel cases and it characterized in that it corresponds to smaller than any positive number when the absolute value of the difference between the amount of change of the radius and the radial pre-set.

Medical images based blood vessel detection device according to the present invention for achieving the above object is a parameter space and generates a parameter space (parameter space) including a blood vessel shown in the medical image obtained through angiography; And variably adjusting the size of the generated spatial parameters, parameterization unit for generating a plurality of curved lines by using the Bezier curve (bezier curve) relative to the normal vector direction is placed into blood vessels; It includes; blood vessel detection unit for selecting at least one curve of the generated plurality of curve.

In addition, the parameterization unit detecting unit for fixing to detect the start and end points of the normal vectors; Part 1 point and the point calculated by repeating the two-point calculated by using the Bezier curve from the surface of the parameter space; And a first curve generator for by the product of the set of points a second set of points to generate a plurality of curve calculation; characterized in that it comprises a.

Medical images based blood vessel detection device according to the above-described present invention, by based on the difference between the brightness value and the spherical coordinate system normal vector brightness values ​​of the radial line in a direction to take advantage of the extraction from the origin of the vascular vessel wall, or general vascular direction or plaque ( plaque) is present and further comprising a blood detection end time determination unit for determining which one of the vessels direction, exit the blood vessel detection time determination unit is the origin of the blood vessel and, inside the average brightness value and a brightness outside the average value of the blood vessel a first extraction unit for extracting; A second extracting section for extracting a plurality of normal vectors facing the direction coordinates of radial lines and brightness relative to the origin of the blood vessel; Vascular direction determination unit for determining the orientation of the blood vessel based on the difference of the inner average brightness value and the brightness value of said radial line; By the radial operation belongs to the set of predetermined conditions for the line, the operation to include a radial straight line corresponding to the direction in which the blood vessel general direction or the plaque exists on the set group set; And satisfied but the condition of the group set is set, the number of elements belonging to the group set set blood vessel detection unit end to end a blood vessel is detected when to converge to zero; may include.

In addition, the vessel direction determination unit brightness of the inner average brightness value and when the positive (+), the difference of the brightness values ​​of the radial line, the radial straight line and said radial straight lines the inner average brightness value determined to be the blood vessel wall, and the difference between the value 0 or a negative (-) when the said radial straight line is characterized in that it is determined that the blood vessel direction under a normal direction or vascular plaque exists.

Further, the operation unit than the absolute value of the difference between the absolute value inside the average brightness of the blood values ​​and the external average brightness value of the difference between the brightness value of the brightness value against the inner average brightness value and the radial line of the radial line is less or equal, and is characterized in that operation in accordance with the group set of the condition set corresponding to the is smaller than any positive number when the absolute value of the difference between the amount of change of the radius and the radial pre-set.

Medical images based blood vessel detection method and apparatus according to the present invention having the configuration as described above is a number by using the medical image obtained through angiography repeat the process of creating a curve based on the normal vector positioned in blood vessel direction by generating the single curve, and select a blood vessel of the generated plurality of curves and the most likely one of the vessel there is an effect that it is possible to detect the exact actual blood vessel.

In addition, methods of medical image based on blood vessel detected by the present invention and the apparatus vessel wall or general vascular direction or plaque and based on the difference of the luminance intensity values ​​of the values ​​and the spherical coordinate system normal to a radial line of the vector direction to take advantage of the extraction from the origin of a blood vessel If the (plaque) is present is determined by any one of the vessels direction, and determines the blood vessel direction under a normal blood vessel direction or plaque exists, the number of elements belonging to the set of conditions preset in accordance with the characteristic that blood vessels are narrower toward the distal end 0 by terminating the detection when the blood vessel to converge, there is an effect that can accurately determine the end point detection vessel.

1 is a view for explaining a configuration of the medical image based on blood vessel detection device according to an embodiment of the invention.

2 is a view for explaining a parametrization employed in the medical image based on blood vessel detection device according to an embodiment of the invention.

3 is a flow chart for explaining the procedure of the medical image based on blood vessel detection method according to an embodiment of the invention.

4 is a view for explaining a first step of the example of the medical image based on blood vessel detection method according to an embodiment of the invention.

5 is a view for explaining the second step of the example of the medical image based on blood vessel detection method according to an embodiment of the invention.

6 is a conceptual view for generating a plurality of curve between the start and end points according to the second step of the medical image based on the blood vessel detection method according to an embodiment of the invention.

7 is a conceptual diagram showing the statistical information extracted from the specific one of the curve in the medical image based on blood vessel detection method according to an embodiment of the invention.

Figure 8 is a schematic diagram showing the average distance to the gray level extracted in the previous step in the medical image based on blood vessel detection method according to an embodiment of the invention.

9 is a view for illustrating an example result of the third step of the medical image based on the blood vessel detection method according to an embodiment of the invention.

10 is a view for explaining a configuration of a blood vessel is detected at the end determining unit according to an embodiment of the invention.

11 is a view for explaining the procedure of a blood vessel is detected at the end making process according to an embodiment of the invention.

12 is a view for explaining a method for extracting a mean luminance value inside and outside of the average brightness value of the blood vessel in the blood vessel is detected at the end making process according to an embodiment of the invention.

13 is a view for explaining a method for extracting the coordinates of a plurality of normal vectors facing the direction relative to the origin of the vessels from the blood vessel is detected at the end making process according to an embodiment of the present invention a radial straight line.

14 is a diagram for explaining the time to end the blood vessels detected in the blood vessel detection end time determination process according to an embodiment of the present invention.

15 is a luminance value of the extracted form ring structures applying the medical image based on blood vessel detection method and the blood vessel is detected at the end making process according to an embodiment of the present invention model.

16 is a plan view of Fig.

17 is an extraction process of a model of a blood vessel tracking operation to extract the model intensity values ​​of FIG.

18 is a view showing an overall construction vessel tracking technique applying the medical image based on blood vessel detection method and the blood vessel is detected at the end making process according to an embodiment of the invention.

Hereinafter to be described in detail enough to easily carry out self technical features of the present invention one of ordinary skill in the art, with reference to the accompanying drawings, the preferred embodiment of the present invention will be described . First, in addition as the reference numerals in the respective drawings of the component, as to the same elements even in different drawings, although the output should be noted that and to have the same reference numerals as much as possible. Further, in the following description of the present invention, a detailed description of known functions and configurations that are determined to obscure the gist of the present invention, the detailed description thereof will be omitted.

Or less, based on medical imaging blood vessel detection method according to an embodiment of the present invention and will be described with respect to the device.

As shown in Figure 1, the medical image based on blood vessel detection device 100 according to the present invention largely comprises a spatial parameter generating unit 110, a parameterization unit 120 and the blood vessel detection section 130.

Referring to Figure 4 parameter space generating section 110 generates the parameter space (parameter space) including a blood vessel shown in the medical image obtained through angiography. The parameters of the virtual space is a cylindrical space that is created is extended by a predetermined distance in vector direction along the blood vessel. Parameter space as shown in Figure 4 can be assumed in a cylindrical shape.

If 4, see FIGS. 5 and 6 with the parameterization unit 120 is a number used for, while variably adjusting the size of the generated parameter space, Bezier relative to the vessels direction of the normal vector curve (bezier curve) It produces a single curve. A normal vector can be seen by the connection 12 to 17 to be described later with FIG.

In this case, the parameterization unit 120 may generate a plurality of curved and variably adjusting the size of the radius (r) of the parameter space. Bezier curve is a curve generated in the cylindrical shape in Fig. Also the curve formed by a manner interconnecting the vector, such as 5, and thus there is a plurality of curves as shown in Figure 6 may be generated.

A normal vector can be formed randomly as a vector toward the end point (e) from the starting point (s) specified in the vessel shown in the medical image obtained through angiography. By generating a vector that can be connected to different points on these points and blood vessel continues repeatedly connected, and appropriate adjustment of the distance between the points to form the curve.

Blood vessel detection unit 130 selects at least one curve of the generated plurality of curve. That is, the blood vessel detection section 130 selects at least one of the plurality of curve to be curve modeling the blood vessel and the most similar. The ever described later curve selection method of a blood vessel detection unit (130).

Such medical image based on blood vessel detection device generates a parameter space (parameter space) including a blood vessel shown in a contrast blood vessel image, Bezier formed while adjusting the size of the parameter space (parameter space) using the curve (bezier curve) to generate a plurality of curves, and by selecting a plurality of curves similar to the curve of the blood vessel and the blood vessel to perform the detection.

In the following, description in more detail for the parameterization unit employed in the medical image based on blood vessel detection device according to an embodiment of the invention.

Establish a horizontal direction in the Z-axis in the following, it jeonghamyeo the Z axis perpendicular to a plane in the XY plane, and the cross-section of the blood vessel in the same direction as the blood vessel at a particular point corresponds to the z-axis, and the vertical plane z-axis can be assumed to xy plane . Wherein θ is the angle of a particular point on the z-axis from the xy plane, φ is the angle in the circumferential direction around the z-axis, ρ is the distance of the z-axis direction.

2, the parameterization unit 120 according to an embodiment of the present invention is used while variably adjusting the size of the generated parameter space, Bezier relative to the normal line vector placed the vessel direction curve (bezier curve) and it may generate a plurality of curves.

I.e. parameterization unit 120 of the number of finite extracts the normal vector of the number of finite to a specific angular interval of θ and φ of the coordinate system obtained for a normal vector specified arbitrarily towards the vessel direction up to 90 degrees and 360 degrees range to the normal vector of each can generate the parameter space and generate a plurality of Bezier curves each.

For this purpose, the parameterization unit 120 includes a detection section 121, a point calculating section 122 and a curve generator (123).

Detector 121 is fixed to detect the start and end points of the normal vectors facing the direction of blood vessels, such as Figs. That is detector 121 is set the start point and the end point of hypothetical normal vector, and may store the respective positions. In FIG. 4, the start and end points of the normal vectors are marked with "s" and "e", is shown in upper case, the Fig.

Figure 5 and the point calculating section 122 as shown in Fig. 6 by calculating the Bezier storage by repeating the first point and the second point from the surface of the parameter space using the curve.

Curve generation part 123 generates a plurality of curves over the product of the set of points a second set of the calculated first point. That is to generate the curve as a curve generating unit 123 is the first point and the number of combinations by each of the combined sequence of two-point calculation.

Hereinafter, a description will be given of the medical image based on blood vessel detection method according to an embodiment of the invention.

Referring to Figure 3, by using the medical image based on blood vessel detection device described medical image based on blood vessel detection method according to the above embodiment of the invention, the description below duplicate will be omitted.

First, create a parameter space (parameter space) including a blood vessel shown in the medical image obtained through angiography in Step 1 (S100). Article as described in the step 1 shown in Figure 4, to produce a cylindrical space of the parameters, including blood vessels.

Next, while adjusting the size of the parameter space produced by the second stage to vary, and generates a plurality of curves and Bezier relative to the normal vector direction is placed in blood vessels using the curve (bezier curve) (S200).

The second step is to fix the starting point (S) and end (E) of the normal vector as shown in (a), (b), (c) and (d) of FIG. Then, by using the Bezier curve is calculated by repeating the first point (p0) and the second point (p1) from the surface of the parameter space.

At this time, the first point and the second and points may be sampled at certain intervals for a certain interval and φ direction to the ρ direction, so that a finite first point and the Bezier curve by calculating a 2 point repeatedly on the parameter space, the can be created.

At this point, the first point to p0, and the second point may be viewed as p1. Then, to generate a plurality of curves over the product of the set of first points in the second set point of the calculation.

At this time, when a set of the first point A d, and referred to a set of the second point B, a Cartesian product AXB is normal vector pair in the set of the second point set and the first point, the number of pairs is soon normal vector a is a number (Nc) of the curve that may exist between the start point (S) and end (E).

Next, select at least one curve of the plurality of curves generated by the first stage 3 (S300). The third step is substantially this iterative implementation results 9 selects at least one of the curves is determined that the previously most similar vessels of the plurality of curves generated in the second step and as shown in Figure 7 and Figure 6, and and it can appear together.

Figure 9 is a left side of the curve, select one of the blood vessels by the black dots is shown by a bold line, and is on the right side is denoted by the points are connected in the parameter space white point.

Thus, the present invention medical image based on blood vessel detection program according to and the device will generate a plurality of curves using the medical image obtained through angiography repeat the process of creating a curve based on the normal vector positioned in blood vessel direction and by selecting the vessel and the most similar to a curve of the generated plurality of curved lines may be detected by a precise actual vascular modeling according to the selected curve.

The single curve most similar to the blood vessel of the plurality of curves in step 3 can be determined by the following equation (1) to 4 same.

Figure 7 shows in the medical image based on blood vessel detection method according to an embodiment of the present invention and the concept showing the statistical information extracted from the specific one of the curve in the medical image based on blood vessel detection method according to an embodiment of the present invention, in Figure 8 concept showing the average distance to the gray level extracted in the previous step are shown.

Equation 1

Figure PCTKR2015000946-appb-M000001

Equation 1 is an equation for calculating the variance for all brightness values ​​lie on the curve.

Here, σ cij refers to the condition that the difference is to be minimized between the sampled values of the points lying on a curve. Substantially veins can be recognized as a blood vessel under the condition that the sampled values ​​of the points lying on the curve having similar values ​​since the leads.

Equation 2

Figure PCTKR2015000946-appb-M000002

Equation 2 is an equation for calculating the average of the distance to the gray level extracted in the previous steps, it can be represented by the equivalent of Fig.

Here, I bijk is a brightness value that is extracted from the sampling points built placed on the curved chopping (ijk b). That is, the I bijk mean brightness value of the k-th sampling point (b ijk) built placed on the curved chopping. t = N t +1 is the number of elements of the set that belong to b ijk. F i is equal to the brightness value of the first (k = 0) b ijk in terms of I bijk brightness value is obtained from the starting point of the track at a particular step.

As equation (2) is a mathematical optimization using the least square error method to the above, as soon as a minimum the deviation established in the equation (1) at the same time that a minimum of the deviation of any of the brightness values, because there may be concluded that it is the vessel, before step the equation is used to select a case as to have a small error as possible brightness values ​​and the luminance values ​​estimated by the average, at the same time the deviations from the minimum.

Equation 3

Figure PCTKR2015000946-appb-M000003

Equation (3) is a formula for picking that the length of the short curve of the curves generated in the multiple parameter space.

Here, b ijk and b ij (k-1) is also represented by a dot, but the other hand, in the area like a vector. The length between the respective vector lies on the curve | b ijk - b ij (k -1) | adding all there is the length of the end curve is such that the deviation is small gray level is about the same as before is to find that the short length of the curve. When the result is obtained as a linear combination of the Equations (1) and (2) each of which is curved to have the same, and a case in which a short length of which can be set at the best curve.

Equation 4

Figure PCTKR2015000946-appb-M000004

In Equation (4) using the parameters of the equations (1) through (3), the energy value of a linear equation that minimizes

Figure PCTKR2015000946-appb-I000001
Find the inner surface of the combinations it can be determined by a curve placed on the vessel.

Here, q i (θ, φ, ρ), s refers to the formula 1, q i (θ, φ , ρ), and d refers to the expression (2) and, q i (θ, φ, ρ), l refers to the expression (3). In addition, w 1, w 2, w 3 means the weight (weighting factor) are respectively preset.

Hereinafter, it will be described with respect to the blood vessel is detected at the end making process and the unit according to the embodiment of possible further advances in methods for detecting the above-described actual vascular invention.

According to the blood vessel is detected at the end making process and the unit, and extract a blood vessel inside the mean gray level and outside the average brightness value, the coordinate values ​​and brightness values ​​of the radial straight line towards the plurality of the normal vector direction based on the vascular origin extraction, and judges the direction of the blood vessel, so that the number of straight line heading in the direction of the blood vessel can be calculated when the blood vessel at the end to converge to zero.

The decision point exit vessel detection by the present invention according to the process and the unit is to, based on the difference between the brightness value and the spherical coordinate system brightness values ​​of a plurality of normal vectors radial straight line in a direction to take advantage of the extraction from the origin of the vascular vessel wall, or general vascular when any one of direction or plaque vascular direction (plaque) the presence of, and possible to determine the normal vector direction, is determined in the normal blood vessel direction or vascular direction plaque is present, the group by the characteristic in which blood vessels are narrower toward the distal end by terminating the blood vessel is detected when the number of elements belonging to a predetermined set of conditions converge to zero, and to accurately determine the end point detection vessel.

10 is a view for explaining a configuration of a blood vessel is detected at the end determining unit according to the invention;

Referring to Figure 10, end blood vessel detection time determination unit 200 according to the present invention is largely the first extraction unit 210 and second extraction unit 220, a blood vessel orientation determining unit 230, a calculation unit (240 ) and a blood vessel detection end portion 250.

A first extraction unit 210 specifies the origin of the blood vessel and extracting the inner average brightness value and the average luminance value outside of the vessel. At this time, the inside of the blood vessel is extracted, and an average brightness value of the internal external relative to the vessel wall to separate them because it is lighter than the contrast blood vessel wall, respectively.

A second extraction unit 220 extracts a plurality of normal vectors facing the direction coordinates of radial lines and brightness relative to the origin of the blood vessel. At this time, the coordinate values ​​of the radial line is used in a spherical coordinate system.

Vascular direction determination unit 230 determines the orientation of the blood vessel based on the difference between the brightness values ​​of the average brightness value and the internal radial lines.

Vascular direction determination unit 230 of the brightness values ​​within the average luminance value and if the difference between the brightness value of a radial straight line positive (+), the radial line is the inner average determined that the blood vessel wall, and brightness value and the radial line the difference is zero or negative (-) if the, the radial line is determined to be the direction in which the blood vessel general direction or plaque exists.

Computing unit 240 computes that in the terms of a predetermined set for the radial straight line, and included in the element group of the set when the set of the radial straight line is determined by the general direction in which the blood vessel or blood vessel plaque direction exists.

That is, the operation unit 240 after the operation that belongs to a group condition of the predetermined set which will be described later to the radial line, the absolute value of the difference between the radial luminance value for the inner average brightness value of the straight line and the radial line brightness vessels for internal average gray level and outside the average is less than the absolute value of the difference of brightness values ​​or the same, and thereby is in the set of predetermined to be smaller than any positive number when the absolute value of the difference between the amount of change of the radius and the radial pre-set. This will be described in detail later.

But satisfies the blood vessel detection end unit 250 may group a set of conditions is set, the group is terminated when the blood vessel detected by the converging to zero the number of elements belonging to a predetermined set.

Referring to Figure 11, by using the blood vessel is detected at the end making process described above blood vessel detection end time determining apparatus according to the present invention, description thereof will be omitted below duplicate.

First, to specify the origin of a blood vessel in one step, and extracts an internal average brightness value and the average brightness value outside of said vessel (S400). The first step is inside such a reference point pi described as detecting the i-th coordinate of the vessel centerline that is in progress and, as shown in Figure 12 based on the blood vessel wall to separate them, because the interior of the blood vessel is bright contrast than the blood vessel wall each lay down the average brightness value by Fi, Bi an average brightness value of the external.

Next, it extracts the number of the normal vectors facing the direction coordinates of radial lines and brightness relative to the origin of a blood vessel in the first stage 2 (S400). At this time, the coordinate values ​​of the radial line is used in a spherical coordinate system. The second step is conducted a plurality of radial straight line towards the plurality of vector direction is Δθ and azimuth Δφ corresponding to the zenith angle forms with the reference point, based on the normal vector by using a coordinate system (θ, φ, ρ) to the extent of Δρ , and it extracts a luminance value as shown in the coordinate value is placed on each straight line in Fig.

Next, on the basis of the difference between the brightness values ​​of the average brightness value and the internal radial straight to step 3 and determines the direction of the blood vessel (S700). The third step is when the difference of brightness values ​​within the average brightness value and the radial line of the positive (+), the radial straight line, and determines that the blood vessel wall, the difference in brightness within the average brightness value and the radial straight line zero or negative, (-) of the case, the radial line is determined to be the direction in which the blood vessel general direction or plaque exists. That is, a is the difference of the luminance values ​​extracted from and the luminance values ​​extracted as the origin, is depending on the sign is positive if the vessel wall, is equal to zero or negative, or the direction of a general blood vessel can be considered in the direction of the blood vessel plaque is present and, for each patient it is the mechanism for responding to plaque occurring irregularly. Therefore going normal vector towards the vessel walls is not included in the elements of the set E.

Next, the calculation by substituting the condition of the set set of radial straight lines as the first stage 4 (S800). Condition of the set used in the fourth step as follows.

That is, when the absolute value of the difference between the brightness value of the brightness value against the inner average brightness value of radial lines and a radial straight line is less than or equal to the absolute value of the difference between the inner average gray level and outside the average brightness value of the blood vessel and radially and It corresponds to the case where the absolute value of the difference between the amount of change in the radius of the group is smaller than any positive number set. At this time, inside the brightness and background brightness of a blood vessel difference is, the vector only blood vessels that are normal vectors higher and with ρ size, and the circulation of the vector does not reach, the blood vessel wall, reaches an Δρ value defined in advance It can be thought of as a vector in the direction progressed. Accordingly, and such that only the vector normal to the guests.

Finally, but satisfies a set of conditions preset in the fifth step is the end of the blood vessel is detected when the number of elements belonging to the group set set converge to 0 (S900). A fifth step in a blood vessel is to say n (E) is the number of elements belonging to the set E using the features to be gradually narrowed to eventually end, the more i progresses n (E) is 0 as shown in Fig. 14 convergence, and the end of the n (E) = 0 when the algorithm. The reason is because there is no longer a normal vector to the brightness value that can move is estimated to vessels that were in progress in the direction ni vessels meet the no.

The method for tracking and detection of the above blood vessel is to go Hannah tracing is performed while the corresponding information (parameter space generated, parametrization, curve creation, selection curve) is repeated. The nagamyeo provide small curve similar to the blood vessels in the process of tracing step by step and fulfill the entire look of one vessel of the curve finally, it can be the core of the vessel detection methods.

In addition, it is to be with the vessel tracking techniques at the end detector is whether laws should stop when satisfied that repetitive behavior is really what conditions. Conversely, the end detection techniques can contribute, so there techniques that can find the end of the vessel, the vessel detection according to vessel tracking techniques to determine whether that should start at some point.

15 is a diagram extracting the brightness values ​​of the ring structure forms applying the medical image based on blood vessel detection method and the blood vessel is detected at the end making process according to an embodiment of the invention the model, and Figure 16 is a plan view of Figure 15, Figure 17 Figure 15 and a gray level extraction models are extracted model of a blood vessel tracking process operation, Figure 18 is a view of the blood vessel tracing technique the whole structure applying the medical image based on blood vessel detection method and the blood vessel is detected at the end making process according to an embodiment of the present invention to be.

According to the blood vessel is detected at the end making process and the unit as shown in Figure 18, if the trace is started from one location to determine whether and when to start or quit the tracking of the blood vessel, and performing step-by-step blood vessel detection, at the same time, the following in step further if you want to repeat the blood is detected, the test is in progress about whether you should stop at that point.

In substantially unpredictable data of the three-dimensional that the vessel in all possible directions in a three-dimensional space of a radial check and perform the blood vessel detection algorithm.

At this time, if applicable, the technical method for detecting a blood vessel as described above it is possible to perform the end point check at the same time by using the direction vector of the radial made in time is also in favor of blood vessel detection.

Thus, two methods will have a significant technological relevance as it is possible, that is, two methods can be applied to the process of determining the blood vessel end detection time proceeds at the same time the blood vessel detection.

Basically, the blood vessel detection method and an internal average brightness value and the number of comparing the brightness values ​​of the radial linear light than if there is no straight line toward the blood vessel orientation (internal average brightness value in the first "start point (reference point) between the blood vessel is detected at the end If the line exists with) and determined to correspond to the blood vessel end <vessel end detection> "," the start point (origin), a straight line is similar to the blood vessel by continuously using the Bezier curve and the parameter space, not a blood vessel at the end as it is assumed that tracks actual vessel <vessel tracking> '.

The origin of the blood vessel is to sense the start of each step in the blood vessel detection. The words used in the aforementioned method for detecting blood vessels and end detecting point somewhat different, but the starting point is the reference point can think of from the point in time steps.

Thus, if in the embodiment of the present invention will perform the blood vessel detection technique (A) corresponds to the blood vessel end technology (B) to shut down the entire process, that is, technology A -> B in a sequential process technology ( "flow" ) it is assumed to be traveling.

Equals shown in Figure 18, the parameters are given by the expressions (1) to 4, the curve can be determined in a blood vessel in a condition that minimizes the parameter values ​​of equation (4). In addition, by using a characteristic that is increasingly narrowing of the blood vessel to the distal end as shown in Figure 18, converge to the n time (E) is to say the number of elements belonging to the set E, the more i progresses n (E) is zero, as described above and the end of the n (E) = 0 when the tracking algorithm.

Has been described in a preferred embodiment according to the invention from above, can be modified in a variety of forms, and those skilled in the art the various changes and modifications without departing from the claims of the present invention it is understood that there can be carried out.

Claims (14)

  1. A first step of creating a parameter space (parameter space) including a blood vessel shown in the medical image obtained through angiography;
    And variably adjusting the size of the generated parameter space, the second step using a Bezier curve (bezier curve) relative to the normal vector direction is placed in blood vessel for generating a plurality of curves; And
    A third step of selecting at least one curve of the generated plurality of curves;
    Medical images based blood vessel detection method characterized in that comprises a.
  2. According to claim 1,
    A first step of creating a parameter space containing the blood vessel shown in the medical image obtained through angiography is
    Medical imaging blood vessels based detection methods, characterized in that to generate the cylindrical space of the parameters, including blood vessels.
  3. According to claim 1,
    The size of the generated parameter space in a second step, while variably adjusted to generate a plurality of curve using a Bezier curve (bezier curve) relative to the normal vector direction is placed in blood vessel,
    Medical images based blood vessel detection method characterized in that the variably adjusting the radial size of the parameter space.
  4. According to claim 1,
    And variably adjusting the size of the generated parameter space, the second step using a Bezier curve (bezier curve) relative to the normal vector direction is placed in blood vessel for generating a plurality of curves,
    The step of securing the start and end points of the normal vectors;
    By using the Bezier curve calculating repeatedly a first point and a second point from the surface of the parameter space; And
    The product of the set of points a second set of the calculated first points to generate a plurality of curves over;
    Medical images based blood vessel detection method characterized in that comprises a.
  5. According to claim 1,
    A third step of selecting at least one curve of the generated plurality of curves,
    Characterized in that selecting at least one of the curves is determined that one of the plurality of curves similar to the blood vessel image based medical vascular detection method.
  6. According to claim 1,
    Blood vessel detection to determine which of the vessels direction based on a difference between the brightness value and the spherical coordinate system normal vector brightness values ​​of the radial line in a direction to take advantage of the extraction from the origin of a blood vessel to the blood vessel wall or general vascular direction or plaque (plaque) present in one exit point further comprises a decision-making process,
    The blood vessel is detected at the end making process,
    A first step of the origin of the blood vessel and extracting the inner average brightness value and the average brightness value outside of said vessel;
    A second step of extracting a plurality of normal vectors facing the direction of the coordinate values ​​and brightness values ​​of the radial line relative to the origin of the blood vessel;
    A third step of determining the direction of the blood vessel based on the difference of the inner average brightness value and the brightness value of said radial line;
    A fourth step of the operation if it is in the set of predetermined conditions with respect to the radial lines, including a radial straight line corresponding to the direction in which the blood vessel general direction or the plaque exists on the set group set; And
    A fifth step of, but satisfies the condition of the group set is set, exit the blood vessel is detected when the number of elements belonging to the group set set converge to zero;
    The medical image-based vessel detection method further comprises.
  7. 7. The method of claim 6,
    In the second step of extracting a coordinate value of a radial straight line towards the plurality of the normal vector direction relative to the origin of the vessels and the brightness value,
    The coordinate values ​​of the radial line is based on medical imaging blood vessel detection method characterized by using the spherical coordinate system.
  8. 7. The method of claim 6,
    In a third step of determining the direction of the blood vessel based on the difference between the brightness values ​​of the average brightness value and the internal radial lines,
    If the difference between the brightness value of the inner average brightness value and the radial line of the positive (+), the radial straight line, and determines that the blood vessel wall, the difference between the brightness value of the inner average brightness value and the radial straight line zero or negative (- ) of the case, the radial line is based on medical imaging blood vessel detection method characterized in that it is determined that the blood vessel direction under a normal direction or vascular plaque exists.
  9. 7. The method of claim 6,
    In the fourth step of the operation if it is in the set of predetermined conditions with respect to the radial lines, including a radial straight line corresponding to the direction in which the blood vessel general direction or the plaque exists on the set group set,
    Wherein the group condition of the set set of the absolute value of the difference between the absolute value inside the average brightness of the blood values ​​and the external average brightness value of the difference between the brightness value of the brightness of a radial straight line value for the inner average brightness value and the radial line than it is, the case and the medical image based on blood vessel detection method characterized in that it corresponds to smaller than any positive number when the absolute value of the difference between the amount of change of the radius and the radial pre-set.
  10. Parameter space and generates a parameter space (parameter space) including a blood vessel shown in the medical image obtained through angiography;
    And variably adjusting the size of the generated spatial parameters, parameterization unit for generating a plurality of curved lines by using the Bezier curve (bezier curve) relative to the normal vector direction is placed into blood vessels;
    Blood vessel detection unit for selecting at least one curve of the generated plurality of curves;
    Medical images based blood vessel detection device comprising: a.
  11. 11. The method of claim 10,
    The parameterization unit,
    Detecting unit for fixing to detect the start and end points of the normal vectors;
    Part 1 point and the point calculated by repeating the two-point calculated by using the Bezier curve from the surface of the parameter space; And
    A set of the calculated first points in the curve generator for generating a plurality of curves over the product of the second set of points;
    Medical images based blood vessel detection device comprising: a.
  12. 11. The method of claim 10,
    Blood vessel detection to determine which of the vessels direction based on a difference between the brightness value and the spherical coordinate system normal vector brightness values ​​of the radial line in a direction to take advantage of the extraction from the origin of a blood vessel to the blood vessel wall or general vascular direction or plaque (plaque) present in one end, and further comprising a time determination unit,
    The blood vessel is detected at the end determining unit,
    The origin of the blood vessel, the first extraction unit for extracting a mean luminance value inside and outside of the average brightness value of the blood vessel;
    A second extracting section for extracting a plurality of normal vectors facing the direction coordinates of radial lines and brightness relative to the origin of the blood vessel;
    Vascular direction determination unit for determining the orientation of the blood vessel based on the difference of the inner average brightness value and the brightness value of said radial line;
    By the radial operation belongs to the set of predetermined conditions for the line, the operation to include a radial straight line corresponding to the direction in which the blood vessel general direction or the plaque exists on the set group set; And
    The group, but satisfies the condition of the predetermined set, blood vessel detection unit end to end a blood vessel is detected when the number of elements belonging to the group set set converge to zero;
    Medical images based blood vessel detection device comprising: a.
  13. 13. The method of claim 12,
    If the vessel direction determination unit difference is positive (+) of the brightness value of the inner average brightness value and the radial straight line, the corresponding radial line is the difference between the brightness value of the judgment, and wherein the inner average brightness value and the radial straight lines to be the blood vessel wall zero or negative (-) in the case, the radial line is based on medical imaging blood vessel detection device characterized in that it is determined that the blood vessel direction under a normal direction or vascular plaque exists.
  14. 13. The method of claim 12,
    The operating section is the brightness of the radial linear value for the inner average brightness value and the absolute value of the difference between the brightness value of a radial straight line is smaller than the absolute value of the difference between the inner average gray level and outside the average brightness value of the blood vessel or equal If and medical image based on blood vessel detection device, characterized in that for computing according to the group set in the condition set corresponding to the is smaller than any positive number when the absolute value of the difference between the amount of change of the radius and the radial pre-set.
PCT/KR2015/000946 2014-10-29 2015-01-29 Method for detecting blood vessels on basis of medical image, and apparatus thereof WO2016068395A1 (en)

Priority Applications (4)

Application Number Priority Date Filing Date Title
KR1020140148520A KR101630231B1 (en) 2014-10-29 2014-10-29 Method and apparatus for detecting vascular based on medical image
KR10-2014-0148520 2014-10-29
KR1020140148531A KR101579902B1 (en) 2014-10-29 2014-10-29 Method and apparatus for determining end point of blood vessel extraction
KR10-2014-0148531 2014-10-29

Publications (1)

Publication Number Publication Date
WO2016068395A1 true WO2016068395A1 (en) 2016-05-06

Family

ID=55857719

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/KR2015/000946 WO2016068395A1 (en) 2014-10-29 2015-01-29 Method for detecting blood vessels on basis of medical image, and apparatus thereof

Country Status (1)

Country Link
WO (1) WO2016068395A1 (en)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6148095A (en) * 1997-09-08 2000-11-14 University Of Iowa Research Foundation Apparatus and method for determining three-dimensional representations of tortuous vessels
JP2008287355A (en) * 2007-05-15 2008-11-27 Sony Corp Registering device, collating device, program, and data structure
KR20110077740A (en) * 2009-12-30 2011-07-07 서울여자대학교 산학협력단 Apparatus and method for vessel and calcification extraction
US20120207366A1 (en) * 2009-10-13 2012-08-16 Agency For Science, Technology And Research Method and system for segmenting a liver object in an image

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6148095A (en) * 1997-09-08 2000-11-14 University Of Iowa Research Foundation Apparatus and method for determining three-dimensional representations of tortuous vessels
JP2008287355A (en) * 2007-05-15 2008-11-27 Sony Corp Registering device, collating device, program, and data structure
US20120207366A1 (en) * 2009-10-13 2012-08-16 Agency For Science, Technology And Research Method and system for segmenting a liver object in an image
KR20110077740A (en) * 2009-12-30 2011-07-07 서울여자대학교 산학협력단 Apparatus and method for vessel and calcification extraction

Similar Documents

Publication Publication Date Title
Rae et al. Recognition of human head orientation based on artificial neural networks
Feddema et al. Weighted selection of image features for resolved rate visual feedback control
Sharp et al. Multiview registration of 3D scenes by minimizing error between coordinate frames
Koch Dynamic 3-D scene analysis through synthesis feedback control
US20030182072A1 (en) Sensor calibration apparatus, sensor calibration method, program, storage medium, information processing method, and information processing apparatus
EP2479726B1 (en) Image comparison system and image comparison method
Waks et al. Cardiac motion simulator for tagged MRI
JP2978406B2 (en) Motion vector field generating apparatus and method according to the elimination of local anomaly
Gramkow On averaging rotations
US20050253870A1 (en) Marker placement information estimating method and information processing device
US20030219147A1 (en) Object tracking apparatus and method
Wei et al. Real-time visual servoing for laparoscopic surgery. Controlling robot motion with color image segmentation
Yoda et al. An automatic wafer inspection system using pipelined image processing techniques
Alpert et al. Improved methods for image registration
CN101299270B (en) Multiple video cameras synchronous quick calibration method in three-dimensional scanning system
US20050018890A1 (en) Segmentation of left ventriculograms using boosted decision trees
JP2004267393A (en) Image processor and image processing method
US20060241412A1 (en) Method for visualizing damage in the myocardium
JP2003058896A (en) Device, method and program for recognizing positioning attitude
CN104008571B (en) Depth camera mannequins acquisition method and network-based Virtual Fitting System
US6594381B2 (en) Computer vision-based technique for objective assessment of material properties in non-rigid objects
US8706281B2 (en) Method and device for producing a master die tool
Ali et al. Registering coronal histological 2-D sections of a rat brain with coronal sections of a 3-D brain atlas using geometric curve invariants and B-spline representation
Wang et al. Multi-camera calibration with one-dimensional object under general motions
Brinkley A flexible, generic model for anatomic shape: Application to interactive two-dimensional medical image segmentation and matching

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15855223

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase in:

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 15855223

Country of ref document: EP

Kind code of ref document: A1