Kurtosis fractional anisotropy, its contrast and estimation by proxy.
Kurtosis fractional anisotropy, its contrast and estimation by proxy.
复制标题
DOI：
10.1038/srep23999
复制
发表时间：
2016-04-04
影响因子：
4.6
通讯作者：
Jespersen SN
中科院分区：
文献类型：
--
作者：
Hansen B;Jespersen SN
通过扩散磁共振成像 (dMRI) 观察到的扩散峰度可能随方向而变化。这种方向变化总结为标量峰度分数各向异性 (KFA)。最近的研究表明，峰度各向异性提供了其他常用 dMRI 标记中不包含的微观结构对比度。我们将 KFA 与固定大鼠大脑和人脑中的其他 dMRI 对比进行比较。然后，我们使用在物理体模中获得的数据以及基于来自体模、大鼠脊髓和人脑的数据的模拟来研究观察到的对比度差异。最后，我们通过评估不同数量的采样方向和信噪比 (SNR) 水平与真实 KFA 的相关性，评估了一种快速估计计算适度的 KFA 代理的策略。我们还映射了该代理的 b 值依赖性。我们发现 KFA 补充了其他 dMRI 指标的对比，特别是分数各向异性 (FA)，它在接近正交的纤维排列中消失，而 KFA 则不会。模拟和虚拟数据支持了这种解释。因此，KFA 是 FA 的补充，可用于评估复杂的组织排列。当沿至少九个方向执行采样且 SNR 较高时，KFA 代理与真实 KFA 密切相关。
The diffusion kurtosis observed with diffusion magnetic resonance imaging (dMRI) may vary with direction. This directional variation is summarized in the scalar kurtosis fractional anisotropy (KFA). Recent studies suggest that kurtosis anisotropy offers microstructural contrast not contained in other commonly used dMRI markers. We compare KFA to other dMRI contrasts in fixed rat brain and in human brain. We then investigate the observed contrast differences using data obtained in a physical phantom and simulations based on data from the phantom, rat spinal cord, and human brain. Lastly, we assess a strategy for rapid estimation of a computationally modest KFA proxy by evaluating its correlation to true KFA for varying number of sampling directions and signal-to-noise ratio (SNR) levels. We also map this proxy’s b-value dependency. We find that KFA supplements the contrast of other dMRI metrics – particularly fractional anisotropy (FA) which vanishes in near orthogonal fiber arrangements where KFA does not. Simulations and phantom data support this interpretation. KFA therefore supplements FA and could be useful for evaluation of complex tissue arrangements. The KFA proxy is strongly correlated to true KFA when sampling is performed along at least nine directions and SNR is high.