MODELING MRI-BASED TISSUE RELAXATION IN THE PRESENCE OF IRON OVERLOAD AND STEATOSIS

项目摘要 铁超载是遗传性的，也可能是通过慢性输血获得的，影响着大约1600万美国人。 脂肪变性(“脂肪过载”)影响三分之一的美国人口，并与肥胖、胰岛素抵抗和 代谢综合征。肝脏铁负荷过重和脂肪变性并存是弥漫性疾病的常见表现 肝病、慢性肝病和癌症治疗，并可引起铁和脂肪毒性，导致 进行性纤维化，不可逆转的肝硬变，最终导致器官衰竭。磁共振成像(MRI)是一种 临床上重要的无创性工具，可独立评估肝脏铁超载和脂肪变性。然而， 在共存条件下，由于铁和脂肪对磁共振成像的混杂影响，MRI定量常常不准确。 信号。已经提出了考虑这些混杂效应的多光谱信号模型 同时量化横向松弛速率(R2*)，预测铁含量和脂肪分数(FF)。 然而，这些模型只在脂肪变性患者中进行了优化和验证，并在不同的联合研究中失败。 存在肝脏铁和脂肪超载的情况。这些模型假定水和脂肪为单R2*或双R2* 质子，以及信号模型中的任何错误假设或不稳定性都会在R2*和Ff中产生误差 计算，导致误诊。使用单R2*还是双R2*的假设取决于退相 体内铁沉积对水和脂肪质子的影响，而这些影响又取决于大小和 铁和脂肪沉积在微观尺度上的分布。先前的模拟和体模研究 研究多光谱信号模型的性能并没有使用真实的组织模型。在模拟中 研究中，没有考虑铁和脂肪分子的大小和分布，而在模型研究中，大小 铁颗粒和脂肪液滴的比例与体内铁和脂肪沉积的比例不匹配。因此，这是一个空白。 在我们理解体内铁和脂肪沉积的真实微观安排将如何导致 磁化率引起的不均匀和影响宏观MRI信号的松弛。在本建议中 研究，我们将进行严格的调查，以评估铁的大小和分布的贡献 和MRI信号上的脂肪沉积通过模拟、模体实验和活体研究来确定 准确的MRI信号模型，可同时、准确地评估铁超载和脂肪变性。我们会 (A)开发一种基于蒙特卡洛的方法，用于创建铁超载、脂肪变性或两者兼有的虚拟肝脏模型 和模拟铁-质子相互作用；(B)构建具有不同颗粒大小的逼真模型，以模拟 体内铁和脂肪沉积；以及(C)通过使用以下方法验证幻影和回顾性患者的MRI信号行为 以活组织检查作为参考标准。这项研究将有助于我们对铁的理解和量化 和脂肪在组织中介导的松弛作用，因此将帮助我们开发和验证准确的信号模型 它可以同时量化R2*和Ff，从而实现对铁过载和 脂肪变性。 1

Project Summary Iron overload, either inherited or acquired through chronic blood transfusions, affects about 16 million Americans. Steatosis (`fat overload') affects one-third of the US population and is linked with obesity, insulin resistance, and metabolic syndromes. Co-occurrence of hepatic iron overload and steatosis is a common manifestation of diffuse liver diseases, chronic hepatopathies, and cancer therapy and can cause iron- and lipo-toxicity leading to progressive fibrosis, irreversible cirrhosis, and ultimately, organ failure. Magnetic resonance imaging (MRI) is a clinically important non-invasive tool for assessing hepatic iron overload and steatosis independently. However, in co-existing conditions, MRI quantification is often inaccurate due to confounding effects of iron and fat on MRI signal. Multi-spectral signal models accounting for these confounding effects have been proposed for simultaneous quantification of transverse relaxation rate (R2*), a predictor for iron content, and fat fraction (FF). However, these models were optimized and validated in only patients with steatosis and failed in different co- existing hepatic iron and fat overload conditions. The models assume either single or dual R2* for water and fat protons, and any incorrect assumptions or instabilities in the signal model produce errors in R2* and FF calculations, leading to misdiagnosis. The assumption to use single or dual R2* depends on the dephasing effects of in vivo iron deposits on water and fat protons, and these effects, in turn, depend on the size and distribution of iron and fat deposits on the microscopic scale. Previous simulation and phantom studies investigating the performances of multi-spectral signal models did not use a realistic tissue model. In simulation study, the sizes and distribution of iron and fat molecules were not considered, and in phantom studies, the sizes of iron particles and fat droplets did not match the scales of in vivo iron and fat deposits. Hence, there is a void in our understanding of how the true microscopic arrangement of iron and fat deposits in vivo will cause susceptibility-induced inhomogeneities and affect the macroscopic MRI signal relaxation. In this proposed research, we will perform a rigorous investigation for evaluating the contribution of size and distribution of iron and fat deposits on MRI signal via simulations, phantom experiments, and in vivo studies to determine an accurate MRI signal model for simultaneous and accurate assessment of iron overload and steatosis. We will (a) develop a Monte Carlo–based approach for creating virtual liver models with iron overload, steatosis, or both and simulating iron-proton interactions; (b) construct realistic phantoms with different particle sizes mimicking in vivo iron and fat deposits; and (c) validate MRI signal behavior in phantoms and retrospective patients by using biopsy assessments as a reference standard. This research will aid our understanding and quantification of iron and fat mediated relaxivity in tissues and therefore, will help us to develop and validate accurate signal models that can simultaneously quantify R2* and FF, thus enabling the noninvasive diagnosis of both iron overload and steatosis. 1