基於前視型血管內超音波系統的血流儲備分數量測：兩種方式比較

Abstract

心血管疾病是現代人常見疾病，在台灣致死率排名第二，其中心肌缺血是此類疾病中相當常見的一種，因冠狀動脈是輸送血液至心臟的關鍵，通常心肌缺血與冠狀動脈疾病關聯性高。會造成心肌缺血的原因眾多，其中冠狀動脈粥狀硬化是常見物理性成因之一，另外成因也可能是低血壓、先天性心臟缺陷、鐮狀細胞性貧血、或因血液異常如癌症引起的血液黏度上升等皆有可能，也因成因眾多，要確認心肌缺血確切原因十分重要但卻困難。血流儲備分數(fraction flow reserve)是目前用來評估心肌缺血是否是由冠狀動脈阻塞造成的黃金指標，然而血流儲備分數測量方式並不方便，傳統的血流儲備分數測量方式需要額外置入血管內壓力感測導絲，不僅需花費額外金錢，且還會因此需要額外手術時間，因此本論文提出利用心血管手術中常使用的血管內超音波系統取代壓力感測導絲進行血流儲備分數測量。為了計算血流儲備分數本論文利用流體力學公式建立壓力損失模型，以白努利(Bernoulli)定律柏達卡諾(Borda-Carnot)方程式與泊肅葉(Poiseuille)定律等公式計算出因阻塞造成的壓力損失，除此之外也利用計算流體力學進行計算，並對利用壓力損失模型與利用計算流體力學兩者進行比較。而為了進行壓力損失模型與計算流體力學需要血管幾何結構與流速，而為了取得這些額外資訊，本論文提出利用訊號處理找出管壁與探頭距離並以此建立血管幾何結構檔案，進行三維幾何重建，並以自相關函數(autocorrelation)取得流速資訊，以進行血流儲備分數計算。本論文以阻塞處出口速度剖面為均勻、為拋物線狀與以入口效應(entrance effect)三種假設建立三種壓力損失模型以及計算流體力學進行壓力損失計算，三種壓力損失模型均方根誤差分別為1.75、2.12與0.72毫米汞柱，而計算流體力學則可以得到0.76毫米汞柱的結果。入口效應模型與計算流體力學備用以血流儲備分數計算，並得到均方根誤差分別為0.033與0.035的結果。最後本論文在血管內超音波仿體實驗中得到均方根誤差3.91毫米汞柱的結果，證明利用血管內超音波搭配流體力學進行血流儲備分數測量的潛力並對其進行探討討論此技術未來改善方向。根據本論文的研究結果，現階段壓力損失模型因含有一難以測得的實驗係數因此難以應用於臨床，但計算流體力學方式則可以提供準確性高且穩定的血流儲備分數計算，在改善血管內超音波的血管幾何結構與血流流速量測準確度後有望代替傳統的壓力導絲量測方式成為新的臨床診斷方式。

Cardiovascular disease is a common illness among modern people and ranks second in Taiwan's mortality rate. Myocardial ischemia is a prevalent type of this disease, and it is often associated with coronary artery disease since the coronary artery is crucial for delivering blood to the heart. Many reasons can cause myocardial ischemia, with coronary artery atherosclerosis being a common physical cause. Other causes may include low blood pressure, congenital heart defects, sickle cell anemia, or abnormalities in the blood, such as an increase in blood viscosity caused by cancer. Due to the numerous possible causes, identifying the exact cause of myocardial ischemia is of crucial importance, but it can also be challenging. Fractional flow reserve (FFR) is currently considered the gold standard for assessing whether myocardial ischemia is caused by coronary artery stenosis. However, the traditional method of FFR measurement could be more convenient, as it requires inserting a pressure-sensing guidewire into the blood vessel. It not only incurs additional costs but also extends the duration of the procedure. Therefore, we propose to use the intravascular ultrasound (IVUS) system commonly used in cardiovascular surgery to replace the pressure sensing guidewire for FFR measurements. To calculate the FFR, we propose to use fluid dynamics formulas to build pressure loss models, including the Bernoulli equation, Borda-Carnot equation, and Poiseuille's law, to calculate the pressure loss caused by stenosis. In addition, computational fluid dynamics (CFD) is also used for calculations and comparisons with pressure loss models. Information on the vascular geometry and flow velocity is required to perform the pressure loss models and CFD calculations. Therefore, we propose signal processing algorithms to find the distance between the vessel wall and the probe and establish a vascular geometry profile for 3D reconstruction. Flow velocity information is obtained using autocorrelation techniques for FFR calculation. Three pressure loss models are established in this study based on three assumptions, including two assuming the flow velocity profile at the outlet of stenosis is uniform and parabolic and the other considering the entrance effect, respectively. The root-mean-square errors for the three pressure loss models are 1.75, 2.12, and 0.72 mmHg, respectively. The CFD results yield a pressure loss of 0.76 mmHg. The entrance effect model and CFD are used for FFR calculation, and the root-mean-square errors obtained are 0.033 and 0.035, respectively. In conclusion, we achieved a root-mean-square error of 3.91 mmHg in the intravascular ultrasound phantom experiments, demonstrating the potential of using intravascular ultrasound combined with fluid dynamics to measure fractional flow reserve and discussing the future direction of improvement for this technique. Based on our research results, the pressure loss model is currently difficult to apply in the clinical setting due to the experimental coefficients that hard to measure, but the computational fluid dynamics method can provide a high level of accuracy and stability in calculating the FFR. After improving the accuracy of vascular geometry reconstruction and flow velocity measurement in intravascular ultrasound, it is expected that CFD method will replace traditional measurement method with pressure guidewire as a new clinical diagnostic method.