The Cardiac Atlas Project

描述（由申请人提供）：心脏畸形是最常见的出生缺陷类型。复杂性先天性心脏病（CHD）治疗的改善导致>90%的先天性CHD患者现在能够存活到成年早期。仅在美国，患有CHD的成年人（约100万人）比儿童多。这些患者中的许多人有心室扩张和功能障碍的风险，特别是那些功能性单心室的患者。预测哪些患者会发生适应不良的重塑以及何时发生是困难的，但医学图像，特别是纵向MRI检查，提供了大量潜在有价值的信息。本项目的目标是使用基于图谱的MR图像衍生的心室形状和生物力学参数模型，识别单心室生理学CHD患者代偿性或适应不良重构的新早期标志物，并通过“心脏图谱项目”（CAP）数据库部署这些模型和数据。CAP是一个全球联盟和在线资源，用于整合和共享心脏成像检查，以及参数模型衍生的功能分析和相关临床信息。我们的儿科心脏病学、医学成像、计算建模和医学信息学专家组成的多学科团队旨在将这一资源扩展到CHD患者，并使用模型来识别适应不良重塑和心力衰竭的早期几何和生物力学预测因子。 该项目将开发计算建模工具，以促进区域心脏形状和机械特征的人群统计分析。模型和相关的本体模式也将促进不同成像协议和模态之间的数据融合。该项目将与加州大学圣地亚哥分校的国家生物计算中心iDASH（整合数据用于分析，分析和共享）合作，并将使用和扩展iDASH开发的数据共享基础设施。 具体目标是：（1）建立60个单心室病变的数据库，包括一些纵向随访。现有的CAP和iDASH基础设施将被扩展，以容纳来自CHD患者的纵向数据，从而能够洞察疾病的进展和失败的发生;（2）识别形状相关性 通过开发基于图谱的心室几何形状、室壁运动和随时间变化的图像导出的统计模型，研究单心室CHD中的机械功能障碍;（3）使用新的基于图谱的CHD模型来实现这些病理中的心脏力学和流体结构相互作用的计算模拟，以导出适应不良重构的潜在早期标记。 这项工作将通过提供区域形状、力学和血流的非侵入性分析对先天性心脏病产生重大影响。如果成功的话，额外的侵入性手术的需求，增加了不良事件的风险，可以大大减少。一个可通过网络访问的数据库也将促进心脏病学、放射学、生理学和计算生物学方面的教育和培训。

DESCRIPTION (provided by applicant): Cardiac malformations are the most common type of birth defect. Improvements in the management of complex congenital heart disease (CHD) have resulted in >90% of those born with CHD now able to survive into early adulthood. In the U.S. alone, there are more adults with CHD (~1 million individuals) than children. Many of these patients are at risk of ventricular dilatation and dysfunction especially those with a functional single ventricle. Predicting those patients who will develop maladaptive remodeling and when is difficult, but there is a wealth of potentially valuable information available in medical images, especially longitudinal MRI exams. The goal of this project is to identify new early markers of compensatory or maladaptive remodeling in CHD patients with single ventricle physiology using atlas-based MR image-derived parametric models of ventricular shape and biomechanics, and to deploy these models and data via the "Cardiac Atlas Project" (CAP) database. CAP is a worldwide consortium and online resource for integrating and sharing cardiac imaging examinations, together with parametric model-derived functional analyses and associated clinical information. Our multi-disciplinary team of experts in pediatric cardiology, medical imaging, computational modeling and medical informatics aims to extend this resource to patients with CHD and use the models to identify early geometric and biomechanical predictors of maladaptive remodeling and heart failure. The project will develop computational modeling tools to facilitate statistical analysis across population groups of regional heart shape and mechanical characteristics. The models and associated ontological schema will also facilitate data fusion between different imaging protocols and modalities. This project will be a collaboration with iDASH (Integrating Data for Analysis, Anonymization, and Sharing), a National Center for Biological Computing at UCSD, and will use and extend the data-sharing infrastructure developed by iDASH. The specific aims are: (1) To create a database of 60 single ventricle pathologies including some with longitudinal follow-ups. Existing CAP and iDASH infrastructure will be extended to accommodate longitudinal data from CHD patients, to enable insights into the progression of disease and the onset of failure; (2) To identify shape correlates of mechanical dysfunction in single ventricle CHD by developing atlas-based image- derived statistical models of ventricular geometry, wall motion and changes over time; (3) To use the new atlas-based models of CHD to implement computational simulations of cardiac mechanics and fluid structure interactions in these pathologies to derive potential early markers of maladaptive remodeling. This work will have a significant impact on congenital cardiology by providing non-invasive analyses of regional shape, mechanics and bloodflow. If successful, the requirement for additional invasive procedures, which have increased risk of adverse events, could be substantially reduced. A web-accessible database will also facilitate education and training in cardiology, radiology, physiology and computational biology.