Modeling heart wall myofibers

了解心壁肌纤维的几何结构是制作个性化心脏功能模型的关键一步，它驱动心脏收缩以泵血。这样的模型可以用于在各种条件和伤害下的治疗计划和指导。为了解决这个问题，我们最近开发了一套医学图像分析算法，旨在从扩散磁共振成像(DMRI)恢复心脏肌纤维的几何形状，扩散磁共振成像是成像纤维组织(如心肌)的最先进技术。我们的主要结果表明，心脏肌纤维处于几何最优的组织中。为了研究这个组织的基本属性，我们开发并验证了从体外人类和动物DMRI数据中提取信息的算法。为了解决体内临床DMRI所面临的挑战，我们改进了这些算法，并展示了它们处理噪声和不完整DMRI采集的潜力。我们建议追求两个目标来进一步发展这项技术：

Understanding the geometrical structure of heart wall myofibers, which drive the contraction of the heart to pump blood, is a crucial step in producing personalized models of heart function. Such models can then be used for therapy planning and guidance in a variety of conditions and injuries. In order to address this issue, we have recently developed a set of medical image analysis algorithms, designed to recover the geometry of the heart myofibers from diffusion magnetic resonance imaging (DMRI), the state-of-the-art technique for imaging fibrous tissue, such as the heart myocardium. Our main results suggest that heart myofibers lie in a geometrically optimal organization. In order to study the basic properties of this organization, we have developed and validated algorithms to extract information from ex vivo human and animal DMRI data. To address challenges faced by in vivo clinical DMRI, we refined these algorithms and demonstrated their potential to handle noisy and incomplete DMRI acquisitions. We propose to pursue two objectives to further develop this technology: