Tissue Characterization with Diffusion MRI

描述（由申请人提供）：在医学影像学中，组织表征是评估脑肿瘤病理和形态的基础。在T1或t2加权MR图像上，即使使用造影剂，病理组织与正常组织的区分往往是有限的。然而，在急性脑卒中的情况下，弥散加权磁共振成像结果非常好。此外，扩散张量成像允许基于主要扩散方向和扩散各向异性的白质束的详细表征。在临床MR系统上进行的弥散加权成像显示，与液体中弥散衰减信号不同，组织中弥散相关的MR信号衰减是非单指数的。发现在肿瘤组织中，这种偏离纯单指数信号衰减的现象特别强烈。一个有充分根据的假设是，这种偏差可能与扩散受限有关，并且它确实可能在微米尺度上提供对细胞结构的见解。最终，在这个分辨率水平上，它是对组织学组织样本的光学显微镜分析，允许肿瘤和正常组织之间的明确区分。本研究的主要目的是通过比较MR扩散测量结果与进行扩散扫描的组织切片中获得的光学显微镜图像数据的详细分析，进一步加深我们对受限扩散的理解。在人类神经纤维样本中，我们想研究扩散各向异性与神经纤维密度或髓鞘形成程度之间的关系。在动物脑肿瘤模型中，我们希望将单指数MR信号衰减的偏差与细胞密度、细胞核与细胞质大小比或细胞大小分布进行比较。磁共振扩散成像广泛应用于医学诊断，并正在迅速探索新的适应症。因此，本研究提出了基础研究，以提高临床MR弥散图像数据的解释。

DESCRIPTION (provided by applicant): In medical imaging, tissue characterization is fundamental for the assessment of brain tumor pathology and morphology. On T1 or T2-weighted MR images, even with contrast agents, distinction of pathologic tissue from normal tissue often is limited. Excellent contrast, however, results from diffusion weighted MR imaging in the case of acute stroke. Moreover, diffusion tensor imaging permits a detailed characterization of white matter tracts based on the principal diffusion direction and diffusion anisotropy. Imaging with very high diffusion weighting, which can be performed on clinical MR systems, has revealed that the diffusion related MR signal decay in tissues, unlike the diffusion attenuated signal in fluids, is non-monoexponential. The departure from a pure monoexponential signal decay was found to be particularly strong in tumor tissue. A well-founded hypothesis is that the deviation may be linked to restricted diffusion and that it may indeed provide insights into cellular architecture at the micrometer dimension level. Ultimately, at this level of resolution, it is light microscopy analysis of histologic tissue samples that permits definitive differentiation between tumor and normal tissue. The main goal of this research is to further our understanding about restricted diffusion by comparing the MR diffusion measurements with detailed analysis of light-microscopy image data obtained in tissue sections that underwent diffusion scans. In human nerve fiber samples, we would like to investigate the relationship between diffusion anisotropy and nerve fiber density or degree of myelination. In animal brain tumor models, we would like to compare the deviation from a monoexponential MR signal decay with cell density, nucleus to cytoplasma size ratio, or cell size distribution. MR diffusion imaging is widely used for medical diagnosis and is rapidly exploring new indications. This study, therefore, proposes basic research, which will improve the interpretation of clinical MR diffusion image data.