Dear Editor The global signal is obtained by averaging the time series over the entire brain.1,2 The global signal is thought to capture background fluctuations common over all brain regions3 and Schölvinck et al. reported the global signal is related to neural activity in anesthetized monkeys.4 This finding provided evidence for a significant neural contribution to the global signal. As well, the correlation between global signals and gray matter signals is very high.5 In a recent study with electroencephalographic (EEG) and functional magnetic resonance imaging (fMRI) signal of resting state in humans, Wong et al. found widespread correlations between the amplitude of the global signal and EEG vigilance in the eyesclosed condition.6 A preprocessing step known as global signal regression (GSR) has been widely used in the restingstate fMRI studies. Prior to the computation of the correlations, a global mean time course is regressed out to remove as much nuisance or “noise” variation as possible. However, it is difficult to remove noise components of the time series data without removing neural components. The application of GSR reduces the impact of motion on functional connectivity measures7 and improves the spatial specificity of functional connectivity maps.5 However, GSR is a very controversial issue in preprocessing as it has been shown that the process can create artifactual negative correlations.3,5,8 Murphy does not support the use of GSR as GSR leads to bellshaped correlation value distributions, centered on zero and can affect connectivity measures.8 Saad et al. analyzed a problem that how do correlation patterns and group differences change after GSR and found that GSR can alter local and longrange correlations, potentially spreading underlying group differences to regions that may never have had any.3 Subsequently, Gotts revisited this issue to argue the problem that GSR could alter restingstate correlations when comparing groups of participants.9 In one of our previous study, the preprocessing steps were completed using AFNI and FSL. In the pipeline, there was a step of removing the global signal. Because of inconsistent conclusion of GSR, we did not regress the global signal in PAGrelated data processing and our reviewers related to the PAG paper also suggested us not to remove the global signal in data processing. In addition, we were so sorry about the figure 3B. We made a mistake. “r=.035” should be taken place with “r=0.35”. If possible, we would likely to have a chance to revised the embarrassing mistake and the revised figure was in the attached. P. Liu1 G. Wang1 Y. Liu1 F. Zeng2 D. Lin3 X. Yang1 F. Liang2 V. D. Calhoun3 W. Qin1 1School of Life Science and Technology, Xidian University, Xi’an, Shaanxi, China 2Acupunture and Tuina School, Chengdu University of Traditional Chinese Medicine, Chengdu, Sichuan, China 3The Mind Research Network, Albuquerque, NM, USA
[1]
K. Kiehl,et al.
Removal of Confounding Effects of Global Signal in Functional MRI Analyses
,
2001,
NeuroImage.
[2]
Rajesh Kumar,et al.
A method for removal of global effects from fMRI time series
,
2004,
NeuroImage.
[3]
Hang Joon Jo,et al.
Trouble at Rest: How Correlation Patterns and Group Differences Become Distorted After Global Signal Regression
,
2012,
Brain Connect..
[4]
M. Schölvinck,et al.
Neural basis of global resting-state fMRI activity
,
2010,
Proceedings of the National Academy of Sciences.
[5]
Simon B. Eickhoff,et al.
An improved framework for confound regression and filtering for control of motion artifact in the preprocessing of resting-state functional connectivity data
,
2013,
NeuroImage.
[6]
Kevin Murphy,et al.
The impact of global signal regression on resting state correlations: Are anti-correlated networks introduced?
,
2009,
NeuroImage.
[7]
M. Fox,et al.
The global signal and observed anticorrelated resting state brain networks.
,
2009,
Journal of neurophysiology.
[8]
Hang Joon Jo,et al.
The perils of global signal regression for group comparisons: a case study of Autism Spectrum Disorders
,
2013,
Front. Hum. Neurosci..
[9]
Thomas T. Liu,et al.
The amplitude of the resting-state fMRI global signal is related to EEG vigilance measures
,
2013,
NeuroImage.