Wang Kaihong, Armour Chlöe H, Guo Baolei, Dong Zhihui, Xu Xiao Yun
Department of Chemical Engineering, Imperial College London, London, UK.
National Heart and Lung Institute, Imperial College London, London, UK.
Int J Numer Method Biomed Eng. 2024 Sep;40(9):e3855. doi: 10.1002/cnm.3855. Epub 2024 Jul 25.
Computational fluid dynamics (CFD) simulations have shown great potentials in cardiovascular disease diagnosis and postoperative assessment. Patient-specific and well-tuned boundary conditions are key to obtaining accurate and reliable hemodynamic results. However, CFD simulations are usually performed under non-patient-specific flow conditions due to the absence of in vivo flow and pressure measurements. This study proposes a new method to overcome this challenge by tuning inlet boundary conditions using data extracted from electrocardiogram (ECG). Five patient-specific geometric models of type B aortic dissection were reconstructed from computed tomography (CT) images. Other available data included stoke volume (SV), ECG, and 4D-flow magnetic resonance imaging (MRI). ECG waveforms were processed to extract patient-specific systole to diastole ratio (SDR). Inlet boundary conditions were defined based on a generic aortic flow waveform tuned using (1) SV only, and (2) with ECG and SV (ECG + SV). 4D-flow MRI derived inlet boundary conditions were also used in patient-specific simulations to provide the gold standard for comparison and validation. Simulations using inlet flow waveform tuned with ECG + SV not only successfully reproduced flow distributions in the descending aorta but also provided accurate prediction of time-averaged wall shear stress (TAWSS) in the primary entry tear (PET) and abdominal regions, as well as maximum pressure difference, ∆P, from the aortic root to the distal false lumen. Compared with simulations with inlet waveform tuned with SV alone, using ECG + SV in the tuning method significantly reduced the error in false lumen ejection fraction at the PET (from 149.1% to 6.2%), reduced errors in TAWSS at the PET (from 54.1% to 5.7%) and in the abdominal region (from 61.3% to 11.1%), and improved ∆P prediction (from 283.1% to 18.8%) However, neither of these inlet waveforms could be used for accurate prediction of TAWSS in the ascending aorta. This study demonstrates the importance of SDR in tailoring inlet flow waveforms for patient-specific hemodynamic simulations. A well-tuned flow waveform is essential for ensuring that the simulation results are patient-specific, thereby enhancing the confidence and fidelity of computational tools in future clinical applications.
计算流体动力学（CFD）模拟在心血管疾病诊断和术后评估中显示出巨大潜力。针对患者且经过良好调整的边界条件是获得准确可靠血流动力学结果的关键。然而，由于缺乏体内血流和压力测量，CFD模拟通常在非针对患者的血流条件下进行。本研究提出一种新方法，通过使用从心电图（ECG）提取的数据调整入口边界条件来克服这一挑战。从计算机断层扫描（CT）图像重建了五个B型主动脉夹层的患者特异性几何模型。其他可用数据包括心搏量（SV）、ECG和四维血流磁共振成像（MRI）。对ECG波形进行处理以提取患者特异性的收缩期与舒张期比率（SDR）。基于使用（1）仅SV以及（2）结合ECG和SV（ECG + SV）调整的通用主动脉血流波形来定义入口边界条件。还将四维血流MRI得出的入口边界条件用于患者特异性模拟，以提供比较和验证的金标准。使用结合ECG + SV调整的入口血流波形进行的模拟不仅成功再现了降主动脉中的血流分布，还准确预测了原发破口（PET）和腹部区域的时间平均壁面剪应力（TAWSS），以及从主动脉根部到远端假腔的最大压差∆P。与仅使用SV调整入口波形的模拟相比，在调整方法中使用ECG + SV显著降低了PET处假腔射血分数的误差（从149.1%降至6.2%），降低了PET处（从54.1%降至5.7%）和腹部区域（从61.3%降至11.1%）TAWSS的误差，并改善了∆P预测（从283.1%降至18.8%）。然而，这些入口波形均无法用于准确预测升主动脉中的TAWSS。本研究证明了SDR在为患者特异性血流动力学模拟定制入口血流波形方面的重要性。良好调整的血流波形对于确保模拟结果具有患者特异性至关重要，从而增强计算工具在未来临床应用中的可信度和逼真度。
