Purpose

Patients with repaired Tetralogy of Fallot (rTOF) will develop dilation of the right ventricle (RV) from chronic pulmonary insufficiency and require pulmonary valve replacement (PVR). Cardiac MRI (cMRI) is used to guide therapy but has limitations in studying novel intracardiac flow parameters. This pilot study aimed to demonstrate feasibility of reconstructing RV motion and simulating intracardiac flow in rTOF patients, exclusively using conventional cMRI and an immersed-boundary method computational fluid dynamic (CFD) solver.

Methods

Four rTOF patients and three normal controls underwent cMRI including 4D flow. 3D RV models were segmented from cMRI images. Feature-tracking software captured RV endocardial contours from cMRI long-axis and short-axis cine stacks. RV motion was reconstructed via diffeomorphic mapping (Deformetrica, deformetrica.org), serving as the domain boundary for CFD. Fully-resolved direct numerical simulations were performed over several cardiac cycles. Intracardiac vorticity, kinetic energy (KE) and turbulent kinetic energy (TKE) was measured. For validation, RV motion was compared to manual tracings, results of KE were compared between CFD and 4D flow.

Results

Diastolic vorticity and TKE in rTOF patients were 4.12 ± 2.42 mJ/L and 115 ± 27/s, compared to 2.96 ± 2.16 mJ/L and 78 ± 45/s in controls. There was good agreement between RV motion and manual tracings. The difference in diastolic KE between CFD and 4D flow by Bland-Altman analysis was − 0.89910 to 2 mJ/mL (95% limits of agreement: − 1.351 × 10⁻² mJ/mL to 1.171 × 10⁻² mJ/mL).

Conclusion

This CFD framework can produce intracardiac flow in rTOF patients. CFD has the potential for predicting the effects of PVR in rTOF patients and improve the clinical indications guided by cMRI.

中文翻译：

---

法洛四联症修复后右心室运动和心内血流的计算模型

目的

修复法洛四联症 (rTOF) 的患者会因慢性肺功能不全而出现右心室 (RV) 扩张，并需要肺动脉瓣置换术 (PVR)。心脏 MRI (cMRI) 用于指导治疗，但在研究新的心内血流参数方面存在局限性。这项试点研究旨在证明仅使用传统 cMRI 和浸没边界法计算流体动力学 (CFD) 求解器重建 RV 运动和模拟 rTOF 患者心内血流的可行性。

方法

四名 rTOF 患者和三名正常对照接受了包括 4D 流的 cMRI。3D RV 模型是从 cMRI 图像中分割出来的。特征跟踪软件从 cMRI 长轴和短轴电影堆栈中捕获 RV 心内膜轮廓。RV 运动通过微分同胚映射 (Deformetrica, deformetrica.org) 重建，用作 CFD 的域边界。在几个心动周期内进行了完全解析的直接数值模拟。测量心内涡量、动能（KE）和湍动能（TKE）。为了验证，将 RV 运动与手动跟踪进行了比较，将 KE 的结果在 CFD 和 4D 流之间进行了比较。

结果

rTOF 患者的舒张期涡度和 TKE 分别为 4.12 ± 2.42 mJ/L 和 115 ± 27/s，而对照组为 2.96 ± 2.16 mJ/L 和 78 ± 45/s。RV 运动和手动追踪之间有很好的一致性。通过 Bland-Altman 分析，CFD 和 4D 流量之间的舒张 KE 差异为 - 0.89910 至 2 mJ/mL（95% 的一致性限制： - 1.351 × 10 ^-2 mJ/mL 至 1.171 × 10 ^-2 mJ/mL）。

结论

该 CFD 框架可以在 rTOF 患者中产生心内血流。CFD 具有预测 PVR 对 rTOF 患者的影响并改善 cMRI 指导的临床适应症的潜力。