Improving hemocompatibility in ventricular assist device therapy using physiological controlstrategies

心力衰竭是发达国家死亡的主要原因之一。尽管近年来心脏辅助装置(VAD)植入的数量迅速增加，心脏移植仍然是终末期心力衰竭的黄金标准治疗方法。然而，移植心脏的数量明显低于现有的需求。在不久的将来，供者心脏的可获得性问题没有预期的解决方案。VAD的基本功能是帮助心脏将血液从左心室泵入主动脉。当今大多数可用的系统都是基于旋转血泵(RBP)的，这种泵通常以恒速或恒流运行。更先进的手术方案旨在使血流适应心血管系统(CVS)的时变需求和心脏不断变化的性能。这需要CVS和RBP之间的相互作用，这可以通过生理控制器来实现。建立VAD系统的主要技术问题之一是最大限度地减少溶血，即由于血液暴露在人工泵送机制下而对血细胞的破坏。在这方面，过去已经致力于优化泵送机构以及传导流动路径的几何形状。到目前为止，这主要是在静态操作条件下完成的，这在生理环境中从未发生过。在拟议的项目中，我们希望研究是否可以通过优化RBP的动态控制来改善血液相容性。为了实现这一点，必须找到一个依赖于操作条件的溶血模型。为此，我们将使用两种替代方法。数据驱动方法，其中映射函数由来自测量数据的参数估计来确定；以及物理驱动方法，其中使用液压系统的计算流体动力学(CFD)模拟。在这些活动的同时，我们将更新我们的实验室混合模拟循环回路，以便于在各种动态、生理和病理负荷条件下进行血液相容性测试。基于模型和测试结果，我们将开发控制算法，将血液相容性作为目标纳入稳健的最优控制问题。此外，还将为Sputnik VAD开发泵流量估计器，以支持动态操作和血液兼容性性能。优化后的控制器将在血液兼容的模拟循环回路中进行动态近生理测试，使用动物血液进行评估。该项目的预期结果是一个集成的泵系统，它不仅可以提供所需的血流动力学，还可以通过使用最优控制策略来减少血液损伤。

Heart failure is one of the main causes of mortality in the developed world. Despite the rapid growth in the number of ventricular assist device (VAD) implantations in recent years, heart transplantation remains the gold standard therapy for terminal stage heart failure. Nevertheless, the number of transplanted hearts is significantly lower than the existing need. There is no expected solution to the problem of donor heart availability in the near future. The basic function of a VAD is to assist the heart in pumping blood from the left ventricle into the aorta. Most of todays available systems are based on rotary blood pumps (RBP) which are typically operated at either constant speed or constant flow. More advanced operating schemes aim at adapting blood flow to the time-variant demand of the cardiovascular system (CVS) and the changing performance of the heart. This requires the interaction between the CVS and the RBP, which can be achieved by a physiological controller. One of the main technical issues in building a VAD system is the minimization of hemolysis, the destruction of blood cells due to the exposure of blood to the artificial pumping mechanism. Much work has been dedicated in the past to optimize the geometry of the pumping mechanism as well as the conducting flow path in this respect. So far this has been mainly done in respect to static operating conditions, which never happens in a physiological context. Within the proposed project, we want to investigate if hemocompatibility can be improved by optimizing the dynamic control of the RBP. In order to achieve this, a model of hemolysis dependent on operating conditions has to be found. For this, we will use two alternative approaches. A data driven approach, where a mapping function is determined by parameter estimation from measurement data and a physics driven approach, where a computational fluid dynamics (CFD) simulation of the hydraulic system is used. In parallel to these activities, we will update our laboratory hybrid mock circulatory loop to facilitate hemocompatibility testing under a wide range of dynamic, physiological and pathological load conditions. Based on models and testing results, we will then develop control algorithms, which incorporate hemocompatibility as objective in a robust optimal control problem. Additionally, a pump flow estimator for the Sputnik VAD will be developed to support the dynamic operation and hemocompatibility performance. The optimized controller will be evaluated in dynamic close-to-physiology-testing in the hemocompatible mock circulatory loop using animal blood. The intended result of this project is an integrated pump system, which can not only provide the required hemodynamics but also decrease blood damage by using an optimal control strategy.