基于并行计算的心脏冠状循环血流仿真研究

Simulation of the Hemodynamic Condition in Coronary Vascular System Based on Parallel Computing

数学建模与仿真是研究心血管系统的重要手段。然而,单机往往无法处理大规模血管网络的建模与仿真,限制了仿真技术在医学研究中的应用。为此应用并行计算方法克服单机的局限性,并在大规模的心脏冠状循环血管系统上测试其可行性。首先,在小型微循环网络上验证了模型的正确性;其次,在人工生成的大规模血管网络上研究了计算集群参数与仿真能力的关系;最后,基于集群仿真了真实的猪心脏冠状循环血管网络中的血流动力学状态。仿真结果显示,并行计算技术能有效地扩展血流动力学的仿真规模,提高仿真效率,并能得到与实验结果一致的仿真结果。

Mathematical modeling and simulation is one of the most important approaches for studying cardiovascular system. However, it is difficult for large-scale vascular networks based on a single computer, which limits the application of the simulation technique in medical research. To overcome the limitations of the single computer, the parallel computing is used to model and simulate the hemodynamic conditions in the large-scale coronary vascular system. Firstly, the hemodynamic model is validated on a small microvascular network. Secondly, the relationship between the parameters of the cluster and the simulating capability is studied on the artificially generated vascular net- works. Finally, the hemodynamic parameters of the porcine coronary vascular system are computed using the cluster. The results show that the parallel computing is able to extend the scale of hemodynamic simulation, improve the computing efficiency, and achieve valid results in comparison to the experimental data.