KR20190139579A - A Method For Detecting Blood Vessel In A Computed Tomography Imagy And An Aparatus Thereof - Google Patents

Abstract

According to an aspect of the present invention, to solve a technical problem, a method for detecting a coronary artery from a medical image performed by an image processing apparatus includes: a step of receiving a medical image including a computed tomography (CT) image; a step of extracting an aortic shape based on the CT image; a step of analyzing the shape of a coronary artery using the extracted aortic shape; and a step of detecting the coronary artery using the analyzed result. According to the present invention, a method and apparatus for detecting a coronary artery in a computed tomography image can detect coronary arteries more easily and accurately than conventional methods.

Description

Method and device for detecting coronary arteries in computed tomography images {A Method For Detecting Blood Vessel In A Computed Tomography Imagy And An Aparatus Thereof}

The present invention is a blood vessel detection method and apparatus based on CTA (Computed Tomography Angiography) image. The present invention relates to a method for easily detecting coronary arteries by using Bayesian Inference in a cardiovascular image obtained by computed tomography and an image processing apparatus thereof.

Cardiovascular diseases are a wide range of diseases including heart disease and vascular disease. Heart diseases include myocardial infarction, hypertension, heart failure, arrhythmia, cardiomyopathy and ischemic due to atherosclerosis progression. Vascular diseases include stroke and peripheral vascular disease.

Cardiovascular disease is a disease with a high mortality rate worldwide, including Korea. Coronary artery disease is an important part of heart disease, including cardiac infarction and angina. Myocardial infarction is the blockage of coronary arteries that supply blood to the heart by atherosclerosis, and angina is the narrowing of the coronary artery.

Accurate analysis of coronary arteries is essential to prevent cardiovascular disease. Coronary arteries stiffen from the heart and are characterized by a thin, long distribution with three-dimensional volume. Therefore, the localization of the major locations of the coronary arteries must be preceded in the analysis.

Accurate detection of coronary arteries has been considered to be a difficult problem, mainly based on tracking and optimization techniques. The follow-up technique has the problem of stopping tracking or tracking the wrong place if the lesion is in the middle of the coronary artery. It is also difficult to apply to complex structures such as branch blood vessels.

An optimization technique is a method of tracking coronary arteries that is expected by mathematical calculations based on the distance between two points. This method can be used to track coronary arteries with high accuracy by carefully calculating sections of coronary arteries. The finer the classification, the higher the computational complexity and the longer the tracking takes.

Tracking techniques are less accurate, and optimization techniques are time consuming. Real clinical environments demand high accuracy in a short time. Therefore, there is a growing need for a new coronary artery detection method and apparatus that can satisfy a certain level of accuracy even in a short time and can easily cope with an urgent emergency.

[Patent Document] Korean Patent Publication No. 10-2015-0061055 "Method and apparatus for detecting lesions of coronary artery from coronary artery CTA image" (Published 2015.06.04.)

The present invention has been made to solve the above-mentioned problems, and an object of the present invention is to provide a method and apparatus for easily and accurately detecting a coronary artery of a heart using a geometric model and Bayesian reasoning. do.

In order to solve the above technical problem, a method for detecting a coronary artery from a medical image performed by an image processing apparatus according to an embodiment of the present invention comprises the steps of: receiving a medical image including a computed tomography (CT) image ; Extracting an aortic shape based on the CT image; Analyzing the shape of the coronary artery using the extracted aortic shape; And detecting the coronary artery using the analysis result.

The extracting of the aorta shape may include rendering the 3D image in the CT image; Detecting the aorta in the image by Maximum intensity projection (MIP); And setting reference axes in the detected aorta.

Analyzing the shape of the coronary artery may include obtaining a blood vessel probability map from the CT image; The method may further include extracting a blood vessel candidate object based on the blood vessel probability map.

The detecting of the coronary artery may further include determining the coronary artery by comparing the blood vessel candidate object extracted in the analyzing the shape of the coronary artery with an object detected by a geometric model and Bayesian Inference. .

In addition, an image processing apparatus according to an embodiment of the present invention for solving the above technical problem includes a receiving unit for receiving a medical image including a computed tomography (CT) image; An extraction unit for extracting an aortic shape based on the CT image; An analysis unit for analyzing the shape of the coronary artery using the extracted aortic shape; A detection unit detecting a coronary artery using the analysis result; And a display unit displaying the coronary artery determined by the detection unit.

The extraction unit for extracting the aorta shape is a rendering unit for rendering a three-dimensional image in the CT image; A detector for detecting an aorta in the image by maximum intensity projection (MIP); And a setting unit for setting reference axes in the detected aorta.

An analysis unit for analyzing the shape of the coronary artery; an acquisition unit for acquiring a blood vessel probability map from the CT image; And an extraction unit for extracting a vascular candidate object based on the vascular probability map.

The detection unit for detecting the coronary arteries further comprises a confirmation unit for determining the coronary arteries by comparing the object detected by the vascular candidate object extracted in the step of analyzing the shape of the coronary artery with a geometrical model and Bayesian Inference (Bayesian Inference) .

According to the method and apparatus for detecting coronary arteries in a computed tomography image according to an exemplary embodiment of the present invention, coronary arteries can be detected more easily and accurately than conventional methods.

In addition, according to the method and apparatus for detecting coronary arteries in a computed tomography image according to an embodiment of the present invention, it is possible to derive accurate results for detecting coronary arteries so that they may be utilized in an actual clinical environment.

1 is a flowchart illustrating a method for detecting a coronary artery according to one embodiment.
2 is an exemplary view for explaining a step of extracting the aortic shape according to an embodiment.
3 is an exemplary view for explaining a method for detecting a coronary artery according to an embodiment.
4 is a qualitative comparison result according to an exemplary embodiment.
5 is a flowchart illustrating an apparatus for detecting a coronary artery according to an embodiment.

DETAILED DESCRIPTION A detailed description of preferred embodiments of the present invention will now be described with reference to the accompanying drawings. It should be noted that the same components in the figures represent the same numerals wherever possible. Specific details are set forth in the following description, which is provided to help a more general understanding of the present invention. In describing the present invention, when it is determined that a detailed description of a related known function or configuration may unnecessarily obscure the subject matter of the present invention, the detailed description thereof will be omitted.

The terminology used herein is to select general terms that are widely used as possible in consideration of the functions of the present invention, but may vary according to the intention or precedent of the person skilled in the art to which the present invention belongs, the emergence of new technologies, etc. have. In addition, in certain cases, there is a term arbitrarily selected by the applicant, and in this case, the meaning will be described in detail in the corresponding description of the present invention. Therefore, the terms used in the present specification should be defined based on the meanings of the terms and the contents throughout the present specification, rather than simple names of the terms.

When a part of the present specification is said to "include" any component, this means that it may further include other components, without excluding the other components unless otherwise stated. In addition, the terms "... unit", "module", etc. described in the specification mean a unit for processing at least one function or operation, which may be implemented in hardware or software or a combination of hardware and software. .

1 is a flowchart illustrating a method for detecting a coronary artery according to one embodiment.

Referring to FIG. 1, a method for detecting a coronary artery according to an embodiment may include receiving a medical image including a computed tomography (CT) image (S110) and extracting an aortic shape based on the CT image. Step S120, analyzing the shape of the coronary artery using the extracted aorta shape (S130) and detecting the coronary artery using the analysis result (S140) may be configured.

Receiving a medical image including the computed tomography (CT) image (S110) includes all images capable of taking a heart to detect a coronary artery, such as a CT image. In one embodiment, the CT image is described as a reference.

The extracting of the aorta shape (S120) may include rendering a three-dimensional image in the CT image, detecting the aorta with maximum intensity projection (MIP) in the image, and the reference in the detected aorta. This can be done by setting the reference axes.

Extracting the aortic shape (S110) is for detecting the position of the aorta as a reference before determining the position of the coronary artery. The aorta is large, thick, and shaped like a cylinder. The cross section has the shape of a large circle and is easy as a reference for determining the position of the coronary artery.

2 is an exemplary view for explaining the step (S120) of extracting the aortic shape.

Image a of Figure 2 is a CT image of the heart rendered in a three-dimensional image. Based on the image a, the heart image taken by CT is rendered as a 3D image to determine the position of the aorta. Image b is a three-dimensional image of the aorta using the maximum intensity projection (MIP) based on the image a.

Image c shows only the aorta as a reference for determining coronary artery location. AA represents the ascending aorta and DA represents the descending aorta. Reference axes r1 and r2 are set in these two aorta, respectively.

The method for setting the reference axis is described in the paper 『B. Jeon et al, Maximum a posteriori estimation method for aorta localization and coronary seed identification, Pattern Recognition 68 (2017) 222-232.

This method takes advantage of the feature that two relationships are close to a circle and have a certain angle and distance. The two aorta can be defined as two references in all x-y planes in three dimensions by r1 and r2 represented linearly as shown in FIG. 2.

3 is an exemplary view for explaining a method for detecting a coronary artery according to an embodiment. Here, the step of analyzing the shape of the coronary artery (S130) will be described.

Analyzing the coronary artery shape may include obtaining a blood vessel probability map from the CT image and extracting a blood vessel candidate object based on the blood vessel probability map.

The blood vessel probability map may be represented by image b of FIG. 3. The vascular probabilities map is derived from a Frangi filter (AF Frangi, WJ Niessen, KL Vincken, MA Viergever, Multiscale vessel enhancement filtering, in: Medical Image Computing and Computer-Assisted Interventation MICCAI98, Springer, 1998, pp. 130-137). Can be obtained using.

A vascular candidate object is extracted by performing a connected component analysis (CCA) from the vascular probability map image. The blood vessel candidate object may be shown in image c of FIG. 3.

The detecting of the coronary artery may include determining a coronary artery by comparing the blood vessel candidate object extracted in the analyzing the shape of the coronary artery with an object detected by a geometric model and Bayesian Inference.

The image d of FIG. 3 represents an object in which coronary artery is expected by applying a geometric model and Bayesian inference. The images e and f of FIG. 3 show coronary artery images confirmed by comparing the blood vessel candidate object of the image c and the object of which the coronary artery of the image d is expected.

Hereinafter, an analysis method using the geometric model and Bayesian reasoning will be described in detail.

The geometric model applied above is shown in Figure 1 below.

그림 1

Pic 1

It is based on the geometric relationship of the heart as a model for the relationship between the reference object r and the target object t. Bayesian reasoning is applied based on the assumption that distances or sizes between major parts of the heart follow a Gaussian distribution.

,

로 각각 표현하였다. 목표물체는 관상동맥이며, 기준물체는 대동맥이다. 여기서 ti 와 rj 는 전처리기법에 따라 복셀 또는 관심영역일 수 있다.Aggregate target candidates and reference objects

,

Represented by respectively. The target object is the coronary artery, and the reference object is the aorta. Here, ti and rj may be voxels or regions of interest according to a preprocessing technique.

Vascular probability map V was obtained from original image I, and vascular progenitor T was obtained by performing CCA to cluster and label V by region.

수식 1

Equation 1

Where I and R are CT images or vascular probability maps and reference objects. P (I) and P (R) are parameters that have a constant value in relation to optimization as constants and are generally excluded from calculation.

으로 표현할 수 있다. 3차원 CT영상을 z축 슬라이스(x-y)로 참조하는 것은 axial view로써 일반적으로 사용하는 진단 view 이다. 수식 1은 심장 CT 영상에서 대동맥의 절단면이 원형에 가까운 모양을 띄는 중요한 특징을 취한 모델링이다.3D object ti is the sum of a number of 2D slices

It can be expressed as Referencing a 3D CT image as a z-axis slice (xy) is a diagnostic view generally used as an axial view. Equation 1 is an important modeling model in which the cut surface of the aorta has a circular shape on the heart CT image.

는 순수 영상데이터에 의해 결정되는 확률 값으로, likelihood에 해당한다.

는 사전에 측정된 기하학적 관계 또는 사전에 정의된 조영제 농도 밝기 값으로써 prior의 역할을 한다. 베이즈추론에서 사후최대확률추정은 수식 2와 같이 2차원 평면을 따라 기준물체 R과 타겟물체T 사이의 기하학적 관계를 추정하여 최종적으로 3차원 목표물체 ti가 추정된다.

Is a probability value determined by pure image data and corresponds to likelihood.

Acts as prior as a previously measured geometric relationship or as a predefined contrast concentration brightness value. In Bayesian reasoning, the posterior maximum probability is estimated by estimating the geometric relationship between the reference object R and the target object T along the two-dimensional plane as shown in Equation 2, and finally the three-dimensional target object ti is estimated.

수식 2

Equation 2

이다. 수식 2는 상기 사후확률이 최대가 될 때 후보물체 ti를 추정한다.here

to be. Equation 2 estimates the candidate object ti when the posterior probability becomes maximum.

Equation 2 based on Bayesian reasoning has two important variables: likelihood and prior. Likelihood prefers objects with the largest possible size, and it is determined that the geometric relationship corresponding to prior is more likely to be coronary artery as the specific probability distribution is closer to the predefined. The probability distributions are assumed to follow a Gaussian distribution.

수식 3

Equation 3

그리고

이다. 수식 3의 likelihood는 해당 값이 0에 가까워질수록 높은 값을 갖는다. here

And

to be. The likelihood of Equation 3 has a higher value as the value approaches zero.

  Geometric-prior expresses the relationship between distance-based reference object and target object as probability.

수식 4

Equation 4

이며 다양한 심장의 크기를 보정하기 위한 스케일 factor이다. 수식 5는 기준물체와 목표물체간의 모든 거리를 고려한 확률 값이다.here

It is a scale factor for correcting various heart sizes. Equation 5 is a probability value considering all distances between the reference object and the target object.

Modeling of contrast brightness corresponding to contrast-prior is as follows.

수식 5

Equation 5

The mean and standard deviation values needed in Equations 4 and 5 are measured in advance from actual CT images, and can be obtained by substituting the equation with reference to Table 1 below, and detecting the i-th target object that satisfies the following equation. The object becomes a coronary artery.

수식 6

Equation 6

표 1

Table 1

5 is a schematic block diagram of an apparatus 200 for detecting a coronary artery in a computed tomography image according to an exemplary embodiment.

As shown in FIG. 5, an apparatus 200 configured to perform a method for detecting a coronary artery in a computed tomography image in accordance with an embodiment of the present invention may include a medical image including a computed tomography (CT) image. Receiving unit for receiving the image, Extraction unit for extracting the aortic shape based on the CT image, Analysis unit for analyzing the shape of the coronary artery using the extracted aortic shape, Detection unit for detecting the coronary artery using the analysis result The display unit may include a display unit displaying the coronary artery determined by the detection unit.

The extraction unit for extracting the aortic shape is a rendering unit for rendering a 3D image in the CT image, a detection unit for detecting the aorta by Maximum intensity projection (MIP) in the image and a reference axis in the detected aorta ( reference axes) can be set.

The analysis unit for analyzing the shape of the coronary arteries may include an acquisition unit for obtaining a vessel probability map in the CT image and an extraction unit for extracting a vascular candidate object based on the vessel probability map.

The detection unit for detecting the coronary arteries may include a confirmation unit for determining the coronary artery by comparing the vascular candidate object extracted in the step of analyzing the shape of the coronary artery with an object detected by a geometric model and Bayesian Inference. have.

The specific functions and operations of the individual components of the coronary artery detection apparatus have already been described above, and thus are omitted in this paragraph.

A control unit (not shown) may be added to the device. The controller may collectively control the receiver 210, the extractor 220, the analyzer 230, the detector 240, and the display 250. For example, the controller may be implemented as a single controller or as a plurality of micro-controllers.

Experimental Results and Analysis

Comparison results according to the quantitative and qualitative experiments according to the present invention are as follows.

Quantitative comparisons compared the detection accuracy of a total of 100 CT images. The subjects to be compared are the commercial software QAngioCT (Medis, https://www.medis.nl/.), Vitrea (Vital, https://www.vitalimages.com/.), And 2D mainly used in research related to the present invention. It is a local geometric method (LGM) technique that utilizes the local characteristics of the base.

For reference, see LG D. Han et al, A fast seed detection using local geometrical feature for automatic tracking of coronary arteries in cta, Computer methods and programs in biomedicine 117 (2) (2014) 495 179-188. .

Detection accuracy is shown in Table 2 below. It can be seen that the detection accuracy of the present invention is the highest as 96%.

표 2

TABLE 2

Qualitative results are shown in FIG. 4. Compared with the results of GT and workstation, branch blood vessels including main blood vessels were detected more and more regions were visualized.

 As can be seen through the descriptions and techniques described above in connection with the present invention, the method and apparatus for detecting coronary arteries in a computed tomography image according to an embodiment of the present invention makes it easier to detect coronary arteries. It can be done correctly.

In addition, according to the method and apparatus for detecting coronary arteries in a computed tomography image according to an embodiment of the present invention, it is possible to derive accurate results for detecting coronary arteries so as to be utilized in an actual clinical environment.

The above-described embodiments of the present invention can be written as a program that can be executed in a computer, and can be implemented in a general-purpose digital computer that operates the program using a computer-readable recording medium.

Computer readable media can be any available media that can be accessed by a computer and includes both volatile and nonvolatile media, removable and non-removable media. In addition, computer readable media may include both computer storage media and communication media. Computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Communication media typically includes computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave, or other transmission mechanism, and includes any information delivery media.

The foregoing description of the present invention is intended for illustration, and it will be understood by those skilled in the art that the present invention may be easily modified in other specific forms without changing the technical spirit or essential features of the present invention. will be. Therefore, it should be understood that the embodiments described above are exemplary in all respects and not restrictive. For example, each component described as a single type may be implemented in a distributed manner, and similarly, components described as distributed may be implemented in a combined form.

The scope of the present invention is shown by the following claims rather than the above description, and all changes or modifications derived from the meaning and scope of the claims and their equivalents should be construed as being included in the scope of the present invention. something to do.

Claims (9)

As a method for detecting a coronary artery in a medical image performed by an image processing apparatus,
Receiving a medical image including a computed tomography (CT) image;
Extracting an aortic shape based on the CT image;
Analyzing the shape of the coronary artery using the extracted aortic shape; And
Detecting the coronary artery using the analysis result
Method for detecting a coronary artery comprising a. The method of claim 1,
Extracting the aortic shape,
Rendering the 3D image in the CT image;
Detecting an aorta in the image by Maximum intensity projection (MIP); And
Setting reference axes in the detected aorta
Method for detecting a coronary artery comprising a. The method of claim 1,
Analyzing the shape of the coronary artery,
Obtaining a blood vessel probability map from the CT image; And
Extracting the blood vessel candidate object based on the blood vessel probability map
Method for detecting a coronary artery comprising a. The method of claim 1,
Detecting the coronary artery,
Confirming the coronary artery by comparing the blood vessel candidate object extracted in the step of analyzing the coronary artery and the object detected by the geometric model and Bayesian Inference
Method for detecting a coronary artery comprising a. A computer-readable recording medium having recorded thereon a program for executing the method of any one of claims 1 to 4. An image processing apparatus configured to perform the method according to any one of claims 1 to 4,
A receiver configured to receive a medical image including a computed tomography (CT) image;
An extraction unit for extracting an aortic shape based on the CT image;
An analysis unit for analyzing the shape of the coronary artery using the extracted aortic shape;
A detection unit detecting a coronary artery using the analysis result; And
A display unit for displaying the coronary artery determined by the detection unit
Device for detecting a coronary artery comprising a. The method of claim 6,
Extraction unit for extracting the aortic shape,
A rendering unit for rendering a 3D image from the CT image;
A detector for detecting an aorta in the image by maximum intensity projection (MIP); And
A setting unit for setting reference axes in the detected aorta
Device for detecting a coronary artery comprising a. The method of claim 6,
Analysis unit for analyzing the shape of the coronary artery,
An acquisition unit for acquiring a blood vessel probability map from the CT image; And
Extraction unit for extracting the blood vessel candidate object based on the blood vessel probability map
Device for detecting a coronary artery comprising a. The method of claim 6,
The detection unit for detecting a coronary artery,
Confirmation unit for determining the coronary artery by comparing the blood vessel candidate object extracted in the step of analyzing the coronary artery and the object detected by the geometric model and Bayesian Inference
Device for detecting a coronary artery comprising a.

Images