Simulation of excitation contraction coupling in ventricular cardiac myocytes

心律失常和纤颤是心源性死亡的主要原因。它们可能是由细胞水平的交替和其他致心律失常过程引起的。Ca2+动力学参与其中。该项目将通过模拟心室心肌细胞的兴奋收缩耦合（ECC）来研究部分已知但远未完全理解的细胞心律失常过程。在细胞内的数千个微小的双裂中，膜兴奋被翻译成Ca2+信号。该问题的大范围空间和时间尺度要求采用多尺度技术，用准静态格林函数描述裂缝中的浓度梯度，并用有限元方法模拟体反应扩散过程。裂缝的离子通道将在分子状态下作为随机过程详细建模。膜电位动力学将是细胞类型和物种特异性的。针对该问题，我们将开发一种混合随机-确定性时间步长管理方法。整体过程的空间和时间尺度范围要求有限元具有空间和时间的自适应能力。我们将开发允许两者同时使用的算法。将采用高阶线性隐式龙格-库塔方法来应对时间步长管理的挑战。问题的规模需要在许多cpu上进行高性能计算，包括适当开发的负载平衡方法。

Arrhythmias and fibrillation are among the prime causes of cardiac death. They may arise from alternans and other arrhythmogenic processes on cell level. Ca2+ dynamics are involved in a variety of them. The project will investigate cellular arrhythmogenic processes, which are known in part but far from completely understood especially in their interaction, by simulating excitation contraction coupling (ECC) in ventricular cardiac myocytes. Membrane excitation is translated into a Ca2+ signal in thousands of tiny dyadic clefts inside the cell. The problem's large range of space and time scales requires a multi-scale technique, which describes the concentration gradients in the clefts by quasi-static Green's functions, and simulates the bulk reaction-diffusion processes with finite element methods (FEMs). The cleft's ion channels will be modeled in molecular state detail as stochastic processes. Membrane potential dynamics will be cell type and species specific. We will develop hybrid stochastic-deterministic time step management specified to the problem. The range of space and time scales in the bulk processes requires spatial and temporal adaptivity of the FEM. We will develop algorithms allowing for the use of both. Higher order linearly implicit Runge-Kutta methods will be applied to meet the time step management challenges. The size of the problem requires high performance computing on many CPUs involving appropriately developed load balancing methods.