Global signal regression
Proposed to account for potential residual artifacts induced by head motion (Power2014)
Global signal contains relevant information about behavior and brain topography (Li2019)
Anticorrelations are biologically relevant and GSR mandates that ~50% of correlations be negative a priori (Psychedelics suppress anticorrelations between brain regions, Loss of consciousness suppresses anticorrelations between brain regions)
Reduces test-retest reliability of graph-theoretical metrics (Andellini2015) and intersession reliability of dynamic brain state properties (Smith2018)
cartographic profile analysis is robust to use of GSR (Shine2016)