MR-assisted PET data optimization for neuroimaging studies

描述（由申请人提供）：正电子发射断层扫描（PET）和磁共振成像（MRI）是目前用于神经学研究的两种最强大的成像模式。最近，能够在人类受试者中同时进行PET和MR数据采集的扫描仪已经成为现实，并且这种新技术开辟了使用顺序采集的数据不可能实现的可能性。一个这样的示例是使用MR数据来改进PET扫描器的性能。虽然PET作为一种技术具有许多优点，包括它可能提供定量手段来评估体内生物过程的事实，但PET测量的准确性受到几个因素的干扰。例如，必须执行衰减和散射校正，以考虑伽马射线光子在到达探测器之前在受试者中的相互作用;必须应用运动校正，以避免由于非自主头部运动而导致的图像退化;由于相对有限的空间分辨率，需要部分容积效应校正;动力学建模需要放射性示踪剂动脉输入功能。的空间 并且同时采集的时间相关的MR数据提供了校正这些混杂效应并提高PET估计的可靠性和再现性的独特机会。虽然许多神经学应用可以从这些方法的改进中受益，但在本提案中，我们将重点放在阿尔茨海默病（AD）上，以证明改进MR-PET定量的潜力。MRI和PET被广泛使用，并在评估AD患者时提供了很大程度上的互补信息。同样重要的是，AD是开发MR-PET的一个很好的测试情况，因为上述混杂因素在该患者人群中特别重要，并且是现有PET研究的实质性限制。具体而言，我们将：（1）开发并验证一种精确的基于MR的头部衰减校正方法。我们假设，使用新的序列成像的骨组织和改进的方法相结合，这些与高分辨率解剖MR，头部衰减图更准确的分割CT获得的，可以获得;（2）提高PET数据的定量使用同时采集的MR数据。我们假设，时间和空间相关的MR数据将允许我们通过执行运动和部分容积效应校正以及估计放射性示踪剂动脉输入函数来提高PET数据的可靠性;（3）评估 MR优化PET测量作为AD疾病进展的生物标志物我们假设MR优化后PET差异的效应量将增加，PET测量的变异性将降低。

DESCRIPTION (provided by applicant): Positron emission tomography (PET) and magnetic resonance imaging (MRI) are two of the most powerful imaging modalities currently in use for neurological studies. Recently, scanners capable of simultaneous PET and MR data acquisition in human subjects have become a reality and this new technology opens up possibilities impossible to realize using sequentially acquired data. One such example is using the MR data for improving the performance of the PET scanner. While PET as a technique has many advantages, including the fact that it could potentially provide a quantitative means to assess in vivo biological processes, the accuracy of the PET measurements is confounded by several factors. For example, attenuation and scatter correction have to be performed to account for the interactions of the gamma-ray photons in the subject before reaching the detectors; motion correction has to be applied to avoid the degradation of the images due to involuntary head movements; partial volume effect correction is required due to the relatively limited spatial resolution; the radiotracer arterial input function is required for kinetic modeling. The spatially and temporally correlated MR data acquired simultaneously offer the unique opportunity to correct for these confounding effects and improve the reliability and reproducibility of the PET estimates. Although many neurological applications could benefit from these methodological improvements, in this proposal we are focusing on Alzheimer's disease (AD) for demonstrating the potential of improved MR-PET quantification. MRI and PET are widely used and provide largely complementary information in assessment of AD patients. Equally important, AD is a great test situation for the development of MR-PET because the confounding factors mentioned above are especially important in this patient population and are a substantial limitation of existing PET research. Specifically, we will: (1) Develop and validate an accurate MR-based head attenuation correction method. We hypothesize that using novel sequences for imaging the bone tissue and improved methods for combining these with high resolution anatomical MR, head attenuation maps more accurate than those obtained from segmented CT can be obtained; (2) Improve the quantification of PET data using the simultaneously acquired MR data. We hypothesize that the temporally and spatially correlated MR data will allow us to improve the reliability of the PET data by performing motion and partial volume effect corrections and estimating the radiotracer arterial input function; (3) Evaluate the added value of MR-optimized PET measurements as biomarkers of disease progression in AD. We hypothesize that the effect size of PET differences would be increased and the variability in PET measurements would be decreased after MR-optimization.