论文部分内容阅读
扩散峭度成像(DKI)引入高阶峭度张量来量化水分子扩散的非高斯程度,并可采用多种扩散敏感因子b值参与模型拟合,能反映组织复杂结构更微小变化,因而在脑神经科学研究与临床诊断中具有独特优势.但目前尚缺乏不同b值组合DKI数据采集对不同脑微结构特征的比较分析和针对不同临床需求制定DKI数据采集的最优b值组合方案的思路.本文从脑脊液、灰质和白质的DKI图像对比度角度,结合基于体素统计分析方法比较了5个不同b值组合与高、低2个b值组合采集方案对DKI成像指标的影响及在胼胝体局部组织微结构特征表达的差异.结果表明,低b值(<1 000 s/mm~2)和高b值(>2 000 s/mm~2)采集成像皆有较强的组织分辨能力,而过多增加b值组合个数会增大拟合误差并增加采集时间;使用高、低2个b值(2 000 s/mm~2与1 000 s/mm~2)的组合采集方案较适合于一般DKI临床诊断.
Diffusion kurtosis imaging (DKI) introduces a higher-order kurtosis tensor to quantify the non-Gaussian distribution of water molecules and can use a variety of diffusion-sensitive factor b values to model fit to reflect more subtle changes in the complex structure of the tissue. Brain neuroscience research and clinical diagnosis has unique advantages.But there is still a lack of different b value combination of DKI data collection of different brain microstructure characteristics of different analysis and clinical needs for the development of DKI data collection optimal b value of the program of ideas In this paper, we compared the DKI images of cerebrospinal fluid (CSF), gray matter and white matter, combined with a voxel-based statistical analysis method, and compared the effects of five different b-value combinations with high and low b- The results showed that both low b values (<1 000 s / mm ~ 2) and high b values (> 2 000 s / mm ~ 2) had strong ability of tissue discrimination Increasing the number of b-value combinations will increase the fitting error and increase the collection time; the combination of two high and low b values (2 000 s / mm 2 and 1 000 s / mm 2) is more suitable In general DKI clinical diagnosis.