Schoch N, Kißler F, Stoll M, Engelhardt S, de Simone R, Wolf I, Bendl R, Heuveline V
Engineering Mathematics and Computing Lab (EMCL), Heidelberg University, Heidelberg, Germany.
Medical Physics in Radiation Oncology, DKFZ Heidelberg, Heidelberg, Germany.
Int J Comput Assist Radiol Surg. 2016 Jun;11(6):1051-9. doi: 10.1007/s11548-016-1397-0. Epub 2016 Apr 12.
Patient-specific biomechanical simulations of the behavior of soft tissue gain importance in current surgery assistance systems as they can provide surgeons with valuable ancillary information for diagnosis and therapy. In this work, we aim at supporting minimally invasive mitral valve reconstruction (MVR) surgery by providing scenario setups for FEM-based soft tissue simulations, which simulate the behavior of the patient-individual mitral valve subject to natural forces during the cardiac cycle after an MVR. However, due to the complexity of these simulations and of their underlying mathematical models, it is difficult for non-engineers to sufficiently understand and adequately interpret all relevant modeling and simulation aspects. In particular, it is challenging to set up such simulations in automated preprocessing workflows such that they are both patient-specific and still maximally comprehensive with respect to the model.
In this paper, we address this issue and present a fully automated chain of preprocessing operators for setting up comprehensive, patient-specific biomechanical models on the basis of patient-individual medical data. These models are suitable for FEM-based MVR surgery simulation. The preprocessing methods are integrated into the framework of the Medical Simulation Markup Language and allow for automated information processing in a data-driven pipeline.
We constructed a workflow for holistic, patient-individual information preprocessing for MVR surgery simulations. In particular, we show how simulation preprocessing can be both fully automated and still patient-specific, when using a series of dedicated MVR data analytics operators. The outcome of our operator chain is visualized in order to help the surgeon understand the model setup.
With this work, we expect to improve the usability of simulation-based MVR surgery assistance, through allowing for fully automated, patient-specific simulation setups. Combined visualization of the biomechanical model setup and of the corresponding surgery simulation results fosters the understandability and transparency of our assistance environment.
软组织行为的患者特异性生物力学模拟在当前手术辅助系统中变得越来越重要，因为它们可以为外科医生提供有价值的辅助诊断和治疗信息。在这项工作中，我们旨在通过为基于有限元法（FEM）的软组织模拟提供场景设置，来支持微创二尖瓣重建（MVR）手术，该模拟可在MVR后的心动周期中模拟患者个体二尖瓣在自然力作用下的行为。然而，由于这些模拟及其基础数学模型的复杂性，非工程师很难充分理解并正确解释所有相关的建模和模拟方面。特别是，在自动化预处理工作流程中设置此类模拟具有挑战性，既要使其针对特定患者，又要在模型方面尽可能全面。
在本文中，我们解决了这个问题，并提出了一个完全自动化的预处理操作符链，用于基于患者个体医学数据建立全面的、针对特定患者的生物力学模型。这些模型适用于基于FEM的MVR手术模拟。预处理方法被集成到医学模拟标记语言的框架中，并允许在数据驱动的管道中进行自动化信息处理。
我们构建了一个用于MVR手术模拟的整体患者个体信息预处理工作流程。特别是，我们展示了在使用一系列专用的MVR数据分析操作符时，模拟预处理如何既能完全自动化又能针对特定患者。我们操作符链的结果进行了可视化，以帮助外科医生理解模型设置。
通过这项工作，我们期望通过允许完全自动化的、针对特定患者的模拟设置来提高基于模拟的MVR手术辅助的可用性。生物力学模型设置和相应手术模拟结果的联合可视化提高了我们辅助环境的可理解性和透明度。
