Quantitative cardiac MRI perfusion for longitudinal studies

描述（由申请人提供）：这项工作旨在开发、评价和使用新的MRI方法，用于心肌灌注储备（MPR）的无创定量评估。这些方法将被设计为对心律失常或ECG信号差的患者有效。这些方法将沿着应用于左心室（LV）和左心房（LA）功能 和LA纤维化测量，以研究由于治疗心房颤动（AF）而引起的变化。具体目标是（1）设计和测试高度欠采样的非门控灌注采集和重建。(2)通过改进处理和使用广泛适用于各种剂量和采集类型的方法识别准确的动脉输入函数，提供稳健的定量灌注估计。(3)确定所提出的定量方法的有效性和可重复性以及在收缩期和舒张末期测量的灌注和MPR的差异，包括确定在定量灌注时收缩期或舒张末期是否最可重复。(4)确定AF消融治疗后灌注储备、心脏功能和左心房纤维化随时间的变化。这将允许确定由于转换为窦性心律而急性改善灌注的程度以及数月后恢复较慢的灌注储备恢复程度。这些目标将通过开发连续采集灌注数据的方法来实现，而无需门控信号。将通过回顾性分析数据完成门控。高级重建方法，包括低秩和时空约束，并采用呼吸和心脏运动的补偿，将开发和测试与快速采集的自门控灌注序列。使用约束数据驱动盲估计技术将获得准确的动脉输入函数。新方法的准确性将通过在AF动物模型中的验证来确定，并通过与动态PET的人体研究进行比较，并对方法的重复性进行表征。对接受消融治疗的患者连续应用这些方法将提供有关心肌重塑相关变化的新信息。开发和使用这种精确和可重复的人体灌注测量将加速临床治疗的评估，并为心脏病患者的管理提供新的工具和知识。

DESCRIPTION (provided by applicant): This work seeks to develop, evaluate and use new MRI methods for non-invasive quantitative assessment of myocardial perfusion reserve (MPR). The methods will be designed to work robustly for patients with arrhythmias or poor ECG signals. The methods will be applied along with left ventricular (LV) and left atrial (LA) function and LA fibrosis measurements to study changes due to treatment for atrial fibrillation (AF). Specific aims are (1) To design and test highly-undersampled ungated perfusion acquisitions and reconstructions. (2) To provide robust quantitative perfusion estimates by improved processing and by identifying an accurate arterial input function using methods that are widely applicable across a range of doses and acquisition types. (3) To determine the validity and repeatability of the proposed quantitative methods and the differences in perfusion and MPR measured at systole and diastole, including determining if systole or diastole is most repeatable when quantifying perfusion. (4) To determine the changes over time in perfusion reserve and cardiac function and left atrial fibrosis, after ablative treatment for AF. This will allow for determining the extent of perfusion improvement acutely due to conversion to sinus rhythm and the degree of recovery of perfusion reserve that recovers more slowly over months. These aims will be accomplished by developing methods to acquire perfusion data continuously, without a gating signal. Gating will be accomplished by analyzing the data retrospectively. Advanced reconstruction methods that include low rank and spatiotemporal constraints, and that employ compensation for respiratory and cardiac motion, will be developed and tested with the rapidly acquired self-gated perfusion sequences. An accurate arterial input function will be obtained using constrained data- driven blind estimation techniques. The accuracy of the new methods will be determined by validation in an animal model of AF, and by comparison to human studies with dynamic PET, and the repeatability of the methods will be characterized. The serial application of the methods to patients undergoing ablation therapy will then provide new information regarding changes associated with myocardial remodeling. The development and use of such accurate and repeatable measurements of perfusion in humans will accelerate evaluation of clinical therapies, and provide new tools and knowledge for the management of patients with cardiac disease.