Motion Corrected Reconstruction for 3D Cardiac Simultaneous PET-MR Imaging: Towards Efficient Assessment of Coronary Artery Disease

冠状动脉疾病(CAD)是西方世界发病率和死亡率的主要单一原因。CAD减少了心肌的血液供应，可能导致胸痛(心绞痛)或心脏病发作。目前CAD诊断是通过各种侵入性和非侵入性测试来执行的。然而，在目前的实践中，不可忽视的数量可能不需要干预的患者涉及侵入性、电离和潜在有害的x射线心导管插入术。磁共振成像(MRI)和正电子发射断层扫描(PET)是两种非常有前途的无创性成像技术，用于冠心病的早期风险评估、指导治疗和治疗监测。这两种技术提供了互补的信息，因此，最近推出的同步PET-MR系统为准确解释PET和MR图像的结果提供了巨大的潜力，并为更好地诊断和治疗CAD提供了新的视角。例如，PET灌注图像与非侵入性冠状动脉MR图像的三维融合可以精确定位导致缺血的狭窄，以指导所需的干预措施。然而，采集过程中不可避免的患者运动(如呼吸和心跳引起的运动)会降低PET和MR图像的图像质量。目前，商用的同步PET-MR扫描仪不具备高效、准确地校正此类运动的技术。目前针对心血管PET-MR运动问题的研究进展主要集中在基于MR信息提高PET图像质量上。此外，这些方法不允许真正同时进行PET和MR采集，导致扫描时间延长，因为诊断性MR图像需要在同时进行PET-MR采集之后采集。原因在于，专用MR采集需要与PET成像同时执行，以便校正PET数据中的运动。这意味着磁共振正被用作一种昂贵的运动矫正设备，限制了其诊断用途。在这项提案中，我们的目标是开发、实施和测试高效的PET-MR采集和重建框架的临床可行性，该框架能够真正同时采集互补的PET和MR诊断信息。我们假设，这可以使用两种模式的协同信息，即功能、解剖和运动MR信息，以及运动校正重建方法中的定量灌注PET数据来实现。提出的方法被认为是临床采用PET-MR心血管成像的重要一步，并最终实现对CAD的有效和准确的非侵入性评估。

Coronary artery disease (CAD) is the leading single cause of morbidity and mortality in the Western world. CAD reduces the blood supply to the cardiac muscle and can lead to chest pain (angina) or heart attack. CAD diagnosis is currently performed by a wide range of invasive and non-invasive tests. However in current practice a non-negligible number of patients that may not need the intervention are referred to invasive, ionizing, and potentially harmful x-ray cardiac catheterization procedures. Magnetic Resonance Imaging (MRI) and Positron Emission Tomography (PET) are two very promising non-invasive imaging technologies for early risk assessment, guidance of therapy and treatment monitoring of CAD. Both technologies provide complementary information, thus the recent introduction of simultaneous PET-MR systems offers great potential for accurate interpretation of findings of PET and MR images and new perspectives for better CAD diagnosis and treatment. For example three-dimensional fusion of PET perfusion images with non-invasive coronary MR images may allow the exact localization of stenosis, causing ischemia, to guide required interventions. However, inevitable patient motion (such as that caused by breathing and heart beating) during the acquisition degrades the image quality of both PET and MR images. Currently, commercial simultaneous PET-MR scanners do not feature technology for efficient and accurate correction of such motion. Current research developments that deal with the motion problem in cardiovascular PET-MR concentrate mainly on improving the quality of PET images based on MR information. Moreover these approaches do not allow truly simultaneous PET and MR acquisitions leading to prolonged scan times, since diagnostic MR images need to be acquired after the simultaneous PET-MR acquisition. The reason is that dedicated MR acquisitions need to be performed concurrently with the PET imaging in order to correct for motion in the PET data. This means that MR is being used as an expensive motion-correction device limiting its diagnostic utility. In this proposal we aim to develop, implement and test the clinical feasibility of an efficient PET-MR acquisition and reconstruction framework that enables truly simultaneous acquisition of complementary PET and MR diagnostic information. We hypothesize that this can be achieved using synergistic information of both modalities i.e. functional, anatomic and motion MR information, and quantitative perfusion PET data in a motion corrected reconstruction approach. The proposed approach is foreseen as an important step towards clinical adoption of PET-MR cardiovascular imaging and lastly towards an efficient and accurate non-invasive assessment of CAD.