Department of Medical Radiation Physics, Lund University, Lund, Sweden.
MAGMA. 2013 Aug;26(4):345-70. doi: 10.1007/s10334-013-0371-x. Epub 2013 Feb 27.
Biophysical models that describe the outcome of white matter diffusion MRI experiments have various degrees of complexity. While the simplest models assume equal-sized and parallel axons, more elaborate ones may include distributions of axon diameters and axonal orientation dispersions. These microstructural features can be inferred from diffusion-weighted signal attenuation curves by solving an inverse problem, validated in several Monte Carlo simulation studies. Model development has been paralleled by microscopy studies of the microstructure of excised and fixed nerves, confirming that axon diameter estimates from diffusion measurements agree with those from microscopy. However, results obtained in vivo are less conclusive. For example, the amount of slowly diffusing water is lower than expected, and the diffusion-encoded signal is apparently insensitive to diffusion time variations, contrary to what may be expected. Recent understandings of the resolution limit in diffusion MRI, the rate of water exchange, and the presence of microscopic axonal undulation and axonal orientation dispersions may, however, explain such apparent contradictions. Knowledge of the effects of biophysical mechanisms on water diffusion in tissue can be used to predict the outcome of diffusion tensor imaging (DTI) and of diffusion kurtosis imaging (DKI) studies. Alterations of DTI or DKI parameters found in studies of pathologies such as ischemic stroke can thus be compared with those predicted by modelling. Observations in agreement with the predictions strengthen the credibility of biophysical models; those in disagreement could provide clues of how to improve them. DKI is particularly suited for this purpose; it is performed using higher b-values than DTI, and thus carries more information about the tissue microstructure. The purpose of this review is to provide an update on the current understanding of how various properties of the tissue microstructure and the rate of water exchange between microenvironments are reflected in diffusion MRI measurements. We focus on the use of biophysical models for extracting tissue-specific parameters from data obtained with single PGSE sequences on clinical MRI scanners, but results obtained with animal MRI scanners are also considered. While modelling of white matter is the central theme, experiments on model systems that highlight important aspects of the biophysical models are also reviewed.
描述磁共振扩散成像实验结果的生物物理模型具有不同程度的复杂性。最简单的模型假设轴突具有相同的大小和方向，而更复杂的模型则可能包括轴突直径和轴突方向离散度的分布。这些微观结构特征可以通过求解反问题从扩散加权信号衰减曲线中推断出来，并且已经在几项蒙特卡罗模拟研究中得到验证。模型的发展与离体和固定神经的微观结构显微镜研究并行，证实了从扩散测量中得出的轴突直径估计值与显微镜测量值一致。然而，在体内获得的结果则不太确定。例如，缓慢扩散的水量低于预期，扩散编码信号显然对扩散时间变化不敏感，这与预期的情况相反。最近对扩散磁共振成像的分辨率限制、水交换率以及微观轴突波动和轴突方向离散度的认识，可能可以解释这些明显的矛盾。了解生物物理机制对组织中水分子扩散的影响，可以用来预测扩散张量成像（DTI）和扩散峰度成像（DKI）研究的结果。因此，可以将缺血性中风等病变研究中发现的 DTI 或 DKI 参数的改变与模型预测的参数进行比较。与预测相符的观察结果可以增强生物物理模型的可信度；与预测不符的观察结果则可以提供如何改进模型的线索。DKI 特别适合于这一目的；它使用比 DTI 更高的 b 值进行测量，因此可以提供更多有关组织微观结构的信息。本文的目的是提供对当前如何在扩散磁共振成像测量中反映组织微观结构的各种特性和微环境之间的水交换率的最新理解的综述。我们重点介绍了使用生物物理模型从临床 MRI 扫描仪上的单个 PGSE 序列获得的数据中提取组织特异性参数的方法，但也考虑了动物 MRI 扫描仪上获得的结果。虽然白质建模是本文的中心主题，但也回顾了针对模型系统的实验，这些实验突出了生物物理模型的重要方面。
