혈관 질환 번역(한국어 원본)1. Introduction
Arterial disease는 cardiology, genetics, clinical medicine, radiology, hemodynamics 등 다양한 분야에서의 접근이 필요하다. 특히, 컴퓨터 기술의 발전과 수치해석 기법의 발전으로 인해 CFD (Computational Fluid Dynamics)나 FSI (Fluid-Structure Intercation) method를 이용한 hemodynamic numerical analysis가 널리 이뤄지고 있다 (Tse et al, 2011; Qian et al, 2010; Marshall et al, 2010; Younis et al, 2004; Valencia et al, 2009; Malvè et al; 2011). Actual patient의 medical imaging를 이용한 numerical model의 경우, CT (Computed Tomography)나 MRI (Magnetic Resonance Imaging), IVUS (Intravascular Ultrasound)와 같은 medical image의 reconstruction을 통해 3차원 혈관 형상을 생성하고 이를 통해 혈류 numerical model을 생성한다 (Goubergrits et al, 2008; Frauenfelder et al, 2007; Vandeghinste et al, 2011; Lee et al, 2007; Kock et al, 2008; Wahle et al, 1999). 그러나 혈관 질환의 특성과 angiography의 어려움으로 인해 대부분의 medical imaging는 환자를 대상으로 이뤄지며 이에 lesion이 발생된 혈류 model을 이용한 연구가 상대적으로 많이 이뤄졌다. 혈관질환의 발생부위 예측이나 기저를 이해하기 위해서는 abnormal 혈관 model 뿐만 아니라 lesion이 발생하기 이전의 혈관 model에 대한 연구가 함께 이뤄져야 한다.
본 연구에서는 LAD에 lesion을 지닌 actual 6명 환자의 CT image를 이용하여 6개의 abnormal arterial model을 생성하였다. 또한 본 연구진이 선행 연구를 통해 제안한 normal arterial model 생성 기술을 이용하여 6명 환자에 대한 6 개의 normal arterial model을 생성하였다 (Ryou et al, 2012). Normal arterial model을 통해 plaque의 발생가능성이 높은 위치를 예측하고 이를 실제 환자의 lesion 발생 위치와 비교하였다. 또한 abnormal arterial model을 통해 lesion이 발생한 이후의 혈류 특성을 규명하였으며 healthy state 혈류와의 차이를 분석하였다. 이를 위해 CFD 계산을 통해 normal 및 abnormal model의 혈류 특성을 비교하였으며 특히, plaque이 발생한 위치의 blood flow나 WSS, OSI 등의 차이를 비교하였다.
2. Methods
2.1. Patient information
실제 환자의 vessel information을 획득하기 위하여 LAD에 lesion을 지닌 6명의 환자의 CT 데이터를 활용하였다. CT system (Lightspeed-pro 16MD, GE)의 scanning paremeters는 다음과 같다. slice acquisition, 0.9 mm; gantry rotation time, 0.27 ms; pitch method, axial prospective; tube current time product, 180 mAs for 50-60 kg and 240 mAs for 70-80 kg; and tube potential, 120 kVp. CT images의 3차원 reconstruction은 3D-DOCTOR (Able Software Corp, ver. 4.0)를 이용하여 수행되었다.
2.2. Normal and abnormal arterial model
혈관 질환은 혈관 전체에서 발생하기 보다는 혈관의 curved area나 bifurcation 등에서 국소적으로 발생하는 것으로 알려져 있다 (Malek et al, 1999). 따라서 normal artery와 lesion이 발생된 abnormal artery의 기하학적 차이는 lesion이 발생된 부위에서 크게 나타난다. Abnormal model을 토대로 normal model을 생성하기 위하여, 본 연구진이 선행 연구에서 제안한 normal arterial model 생성 방법을 이용하였다 (Ryou et al, 2012). abnormal 혈관의 plaque을 가상적으로 제거함으로써 정상 상태 혈관을 생성 할 수 있다고 가정했다 (Knight et al, 2010; Olgac et al, 2009). CT image에서 plaque 및 lumen 영역의 contrast 차이와 boundary line 수정 작업을 통해 initial normal model을 생성하였으며 이를 VH-IVUS (Virtual Histology Intravascular Ultrasound)에서 제공한 혈관 단면적 정보를 통해 검토 및 수정하였다 (Ryou et al, 2012). Figure 1. 은 본 연구에서 적용한 normal model을 생성하는 절차를 설명하는 그림이며 Figure 2. 은 12 cases 의 혈류 수치 model이다. abnormal 혈관 model의 plaque을 가상적으로 제거하여 plaque가 제거된 normal arterial model의 형상을 확인할 수 있다.
2.3. Computational simulation
CFD technique를 이용하여 6명의 실제 환자의 normal 및 abnormal arterial model(total 12 cases)의 혈류 특성을 규명하였다. The blood는 lamina, unsteady, incompressible 유동으로 간주하였다. blood와 같은 non-Newtonian fluid의 shear-thinning phenomenon을 모사하기 위하여 Carreau viscosity model을 이용하였다. 이는 shear rate tensor의 second invariant로 표현되며 blood flow의 rheological characteristics 표현에 적합한 것으로 알려져 있다. (Ro et al, 2008; Cho et al, 1985). |
혈관 질환 번역(영어 번역본)1. Introduction
Arterial disease must be approached in various fields like cardiology, genetics, clinical medicine, radiology, hemodynamics, etc. Especially, hemodynamic numerical analysis using CFD (Computational Fluid Dynamics) and FSI (Fluid-Structure Interaction) methods are most widely used due to the development of computer technology and numerical analysis. (Tse et al, 2011; Qian et al, 2010; Marshall et al, 2010; Younis et al, 2004; Valencia et al, 2009; Malvè et al; 2011). In the case of numerical models using medical imaging of actual patients, the three-dimensional blood vessel form is created through the reconstruction of medical images like CT (Computed Tomography), MRI (Magnetic Resonance Imaging), and IVUS (Intravascular Ultrasound), and the numerical model of the blood vessels are created (Goubergrits et al, 2008; Frauenfelder et al, 2007; Vandeghinste et al, 2011; Lee et al, 2007; Kock et al, 2008; Wahle et al, 1999). However, due to the characteristics of arterial diseases and difficulties of angiography, most of the medical imaging is conducted on patients and there are relatively many studies using the blood vessel model with lesion. There must be studies on not only abnormal arterial models but also on arterial models prior to the occurrence of the lesion to understand the prediction or basis of the locations where arterial diseases are caused.
This study created six abnormal arterial models using CT images of actual six patients with lesions in the LAD. In addition, six normal arterial models of the six patients were created using the normal arterial model creation technology that was suggested in previous studies conducted by the researchers of this study (Ryou et al, 2012). We predicted locations where there was a high possibility of plaque occurrence and compared it to the actual locations of occurrence in each patient. In addition, we investigated the flow characteristics after lesion occurred through the abnormal arterial model and compared it to the blood flow in the healthy state. For this, we compared the flow characteristics of normal and abnormal models through CFD calculations, especially concentrating on the blood flow of the locations of plaque occurrence or differences in WSS, OSI.
2. Methods
2.1. Patient Information
We used the CT data of six patients with lesions in LAD in order to achieve the vessel information of actual patients. The scanning parameters of the CT system (Light speed-pro 16MD, GE) are as the following. slice acquisition, 0.9 mm; gantry rotation time, 0.27 ms; pitch method, axial prospective; tube current time product, 180 mAs for 50-60 kg and 240 mAs for 70-80 kg; and tube potential, 120 kVp. The reconstruction of three-dimensional CT images was conducted using 3D-DOCTOR (Able Software Corp, ver. 4.0).
2.2 Normal and Abnormal Arterial Model
Arterial diseases are known to occur mostly in the curved areas or bifurcations of blood vessels rather than throughout the whole blood vessel (Malek et al, 1999). Therefore, the geometric difference between normal arteries and abnormal arteries with lesion is great in the parts where lesion has been created. In order to create a normal model based on the abnormal model, the normal arterial model production method that was suggested by our researchers in a previous study was used (Ryou et al, 2012). It was assumed that by hypothetically removing the plaque of the abnormal artery, we can create a normal artery (Knight et al, 2010; Olgac et al, 2009). The initial normal model was created by manually correcting the boundary lines and the contrast difference of the plaque or lumen area of the CT image. This initial normal model was then checked and corrected by the artery’s cross section information provided by the VH-IVUS (Virtual Histology Intravascular Ultrasound) (Ryou et al, 2012). Figure 1 is a figure explaining the procedure of creating the normal model that was applied in this study and Figure 2 is the blood flow numerical data model of the twelve cases. The form of the normal arterial model without a plaque can be checked by hypothetically removing the plaque of the abnormal artery model.
2.3. Computational Simulation
The flow characteristics of the normal and abnormal arterial model (total 12 cases) of six actual patients were investigated using the CFD technique. The blood was regarded to be lamina, unsteady, incompressible. The Carreau viscosity model was used to describe the shear-thinning phenomenon of a non-Newtonian fluid like blood. This is described as the second invariant of a shear rate tensor that is appropriate for describing the rheological characteristics of blood flow. (Ro et al, 2008; Cho et al, 1985). |
