Anticorrelations

Submitted by alex on Sat, 05/23/2009 - 21:22

 Hi again,
I am interested in using REST to examine anticorrelations. I just wanted to check the following:
1 - If I include a global signal timeseries as a covariable, then the analysis will produce maps showing voxels significantly correlated with the seed timecourse after controlling for global signal fluctuations. That is, it is essentially like performing a global signal regression. Is this correct?
2 - Anticorrelations will be revealed by negative values in the resulting connectivity and z-maps. Is this correct?

thanks for your help,
Alex

Submitted by YAN Chao-Gan on Sun, 05/24/2009 - 15:10

In the past few years, there has been increased attention to the anticorrelation phenomenon of resting fMRI. A typical case is that, while the global (whole-brain) signal was removed, many researchers consistently observed that there were significant anti-correlations between the components of the default-mode and attention networks [14,15,45]. Recently, the global signal has been found to be associated with respiration-induced fMRI signal [40]. To reduce the effect of the physiological artifacts, we therefore removed the whole-brain signal by a regression analysis as shown in previous studies. Of note, it is still an ongoing controversy question about the removal of the global brain signal since the removal causes the re-distribution of correlation coefficients and the ambiguous interpretation of biological mechanisms of correlations [50].

14. Fox MD, Snyder AZ, Vincent JL, Corbetta M, Van Essen DC, et al. (2005) The human brain is intrinsically organized into dynamic, anticorrelated functional networks. Proc Natl Acad Sci U S A 102: 9673-9678.
15. Fransson P (2005) Spontaneous low-frequency BOLD signal fluctuations: an fMRI investigation of the resting-state default mode of brain function hypothesis. Hum Brain Mapp 26: 15-29.
45. Buckner RL, Andrews-Hanna JR, Schacter DL (2008) The Brain's Default Network: Anatomy, Function, and Relevance to Disease. Ann NY Acad Sci 1124: 1-38.
40. Birn RM, Diamond JB, Smith MA, Bandettini PA (2006) Separating respiratory-variation-related fluctuations from neuronal-activity-related fluctuations in fMRI. Neuroimage 31: 1536-1548.
50. Murphy K, Birn RM, Handwerker DA, Jones TB, Bandettini PA (2009) The impact of global signal regression on resting state correlations: are anti-correlated networks introduced? Neuroimage 44: 893-905.

It's a complex problem of global signal removal. For your questions:
1.    Yes. You can understand as that. The calculating is like what AFNI does. Xiao-Wei could give more specific comments.
2.    Yes. Negative values in FCMap or zFCMap means anticorrelations.