Montreal Neurological Institute, Brain Imaging Center, McGill University, Canada.
Neuroimage. 2013 Nov 15;82:393-402. doi: 10.1016/j.neuroimage.2013.05.065. Epub 2013 May 26.
Cross-sectional analysis of longitudinal anatomical magnetic resonance imaging (MRI) data may be suboptimal as each dataset is analyzed independently. In this study, we evaluate how much variability can be reduced by analyzing structural volume changes in longitudinal data using longitudinal analysis. We propose a two-part pipeline that consists of longitudinal registration and longitudinal classification. The longitudinal registration step includes the creation of subject-specific linear and nonlinear templates that are then registered to a population template. The longitudinal classification step comprises a four-dimensional expectation-maximization algorithm, using a priori classes computed by averaging the tissue classes of all time points obtained cross-sectionally. To study the impact of these two steps, we apply the framework completely ("LL method": Longitudinal registration and Longitudinal classification) and partially ("LC method": Longitudinal registration and Cross-sectional classification) and compare these with a standard cross-sectional framework ("CC method": Cross-sectional registration and Cross-sectional classification). The three methods are applied to (1) a scan-rescan database to analyze reliability and (2) the NIH pediatric population to compare gray matter growth trajectories evaluated with a linear mixed model. The LL method, and the LC method to a lesser extent, significantly reduced the variability in the measurements in the scan-rescan study and gave the best-fitted gray matter growth model with the NIH pediatric MRI database. The results confirm that both steps of the longitudinal framework reduce variability and improve accuracy in comparison with the cross-sectional framework, with longitudinal classification yielding the greatest impact. Using the improved method to analyze longitudinal data, we study the growth trajectories of anatomical brain structures in childhood using the NIH pediatric MRI database. We report age- and gender-related growth trajectories of specific regions of the brain during childhood that could be used as a reference in studying the impact of neurological disorders on brain development.
横断面分析纵向解剖磁共振成像（MRI）数据可能不是最优的，因为每个数据集都是独立分析的。在这项研究中，我们通过使用纵向分析来评估在纵向数据中分析结构体积变化可以减少多少可变性。我们提出了一个两部分的管道，包括纵向配准和纵向分类。纵向配准步骤包括创建特定于主体的线性和非线性模板，然后将其注册到人群模板。纵向分类步骤包括使用四阶期望最大化算法，使用通过对所有时间点的组织类别进行平均计算得出的先验类别进行分类。为了研究这两个步骤的影响，我们完全应用该框架（“LL 方法”：纵向配准和纵向分类）和部分应用（“LC 方法”：纵向配准和横断面分类），并将其与标准的横断面框架（“CC 方法”：横断面配准和横断面分类）进行比较。将这三种方法应用于（1）扫描-再扫描数据库以分析可靠性，（2）NIH 儿科人群以比较使用线性混合模型评估的灰质生长轨迹。LL 方法，以及在较小程度上的 LC 方法，在扫描-再扫描研究中显著降低了测量的可变性，并为 NIH 儿科 MRI 数据库提供了最佳拟合的灰质生长模型。结果证实，与横断面框架相比，纵向框架的两个步骤都可以降低变异性并提高准确性，而纵向分类的影响最大。使用改进的方法分析纵向数据，我们使用 NIH 儿科 MRI 数据库研究儿童期解剖大脑结构的生长轨迹。我们报告了儿童期大脑特定区域的与年龄和性别相关的生长轨迹，这些轨迹可用于研究神经疾病对大脑发育的影响。
