The Turner Institute for Brain and Mental Health, School of Psychological Sciences, Monash Biomedical Imaging, Monash University, Victoria, 3168, Australia.
School of Physics, The University of Sydney, NSW, 2006, Australia.
Neuroimage. 2020 May 15;212:116614. doi: 10.1016/j.neuroimage.2020.116614. Epub 2020 Feb 19.
One of the most controversial procedures in the analysis of resting-state functional magnetic resonance imaging (rsfMRI) data is global signal regression (GSR): the removal, via linear regression, of the mean signal averaged over the entire brain. On one hand, the global mean signal contains variance associated with respiratory, scanner-, and motion-related artifacts, and its removal via GSR improves various quality-control metrics, enhances the anatomical specificity of functional-connectivity patterns, and can increase the behavioral variance explained by such patterns. On the other hand, GSR alters the distribution of regional signal correlations in the brain, can induce artifactual anticorrelations, may remove real neural signal, and can distort case-control comparisons of functional-connectivity measures. Global signal fluctuations can be identified visually from a matrix of colour-coded signal intensities, called a carpet plot, in which rows represent voxels and columns represent time. Prior to GSR, large, periodic bands of coherent signal changes that affect most of the brain are often apparent; after GSR, these apparently global changes are greatly diminished. Here, using three independent datasets, we show that reordering carpet plots to emphasize cluster structure in the data reveals a greater diversity of spatially widespread signal deflections (WSDs) than previously thought. Their precise form varies across time and participants, and GSR is only effective in removing specific kinds of WSDs. We present an alternative, iterative correction method called Diffuse Cluster Estimation and Regression (DiCER), that identifies representative signals associated with large clusters of coherent voxels. DiCER is more effective than GSR at removing diverse WSDs as visualized in carpet plots, reduces correlations between functional connectivity and head-motion estimates, reduces inter-individual variability in global correlation structure, and results in comparable or improved identification of canonical functional-connectivity networks. Using task fMRI data across 47 contrasts from 7 tasks in the Human Connectome Project, we also present evidence that DiCER is more successful than GSR in preserving the spatial structure of expected task-related activation patterns. Our findings indicate that care must be exercised when examining WSDs (and their possible removal) in rsfMRI data, and that DiCER is a viable alternative to GSR for removing anatomically widespread and temporally coherent signals. All code for implementing DiCER and replicating our results is available at https://github.com/BMHLab/DiCER.
静息态功能磁共振成像（rsfMRI）数据分析中最具争议的程序之一是全局信号回归（GSR）：通过线性回归去除整个大脑平均信号的均值。一方面，全局平均信号包含与呼吸、扫描仪和运动相关伪影相关的方差，通过 GSR 去除可以提高各种质量控制指标，增强功能连接模式的解剖特异性，并可以增加此类模式解释的行为方差。另一方面，GSR 改变了大脑中区域信号相关性的分布，可能会产生人为的负相关，可能会去除真实的神经信号，并可能扭曲功能连接测量的病例对照比较。全局信号波动可以从称为地毯图的彩色编码信号强度矩阵中进行视觉识别，其中行表示体素，列表示时间。在 GSR 之前，通常会出现影响大脑大部分区域的大周期性相干信号变化带；在 GSR 之后，这些明显的全局变化大大减少。在这里，我们使用三个独立的数据集表明，重新排列地毯图以强调数据中的聚类结构可以揭示比以前认为的更多样化的空间广泛信号偏转（WSD）。它们的精确形式随时间和参与者而变化，并且 GSR 仅能有效去除特定类型的 WSD。我们提出了一种替代的迭代校正方法，称为弥散聚类估计和回归（DiCER），它可以识别与大相干体素簇相关的代表性信号。与 GSR 相比，DiCER 在地毯图中更有效地去除各种 WSD，减少功能连接与头部运动估计之间的相关性，减少全局相关结构中的个体间变异性，并在识别典型功能连接网络方面产生可比或更好的结果。使用来自人类连接组计划中的 7 个任务的 47 个对比的任务 fMRI 数据，我们还提供了证据表明，DiCER 在保留预期任务相关激活模式的空间结构方面比 GSR 更成功。我们的研究结果表明，在 rsfMRI 数据中检查 WSD（及其可能的去除）时必须谨慎行事，并且 DiCER 是 GSR 去除解剖上广泛且时间上相干信号的可行替代方法。所有用于实现 DiCER 和复制我们结果的代码都可在 https://github.com/BMHLab/DiCER 上获得。
