, 2003), signal envelope correlations buy Ixazomib were particularly robust for high-frequency activity. Gamma-band power envelope correlations have also been reported for sensorimotor networks (He et al., 2008), which were low for slow-wave sleep but high for REM sleep and awake state. Task-related decreases in gamma-band power have been demonstrated in the default-mode network (Jerbi et al., 2010). In addition, anticorrelated gamma-band power fluctuations for different

networks have been observed in invasive human recordings (Keller et al., 2013). Several EEG studies have suggested that the dynamics of the slow fluctuations giving rise to envelope ICMs may be scale-free, that is, not characterized by defined peaks in the power spectrum (Linkenkaer-Hansen et al., 2001, He et al., 2010 and Palva and Palva, 2011). Only recently, a number of studies have aimed to investigate the neurophysiology of ICMs by combining noninvasive MEG recordings with source space analyses. Several of these studies used amplitude or see more power envelope correlations (de Pasquale et al., 2010, Brookes et al., 2011, Brookes et al., 2012, Hipp et al., 2012 and de

Pasquale et al., 2012), while others employed phase coherence (Hipp et al., 2011 and Bardouille and Boe, 2012), phase lag index (Hillebrand et al., 2012), or imaginary coherence (Marzetti et al., 2013). An interesting result is that plain correlation of signal envelopes yields spatially unspecific correlation patterns characterized by high correlation of the seed with neighboring voxels and a monotonic drop off to more distant sites (Hipp et al., 2012). While also comprising true interactions, such patterns are likely to reflect, to a substantial amount, spurious correlations arising from volume spread (Nolte et al., 2004 and Hipp

et al., 2012). However, ICM dynamics can be recovered if correlation patterns resulting from volume conduction are suppressed before analyzing functional connectivity (Hipp et al., 2012, Brookes else et al., 2012, Hillebrand et al., 2012 and Marzetti et al., 2013). A recent study that successfully employed this approach for investigation of envelope ICMs has used phase orthogonalization (Figure 2B) to remove zero-phase coupling (Hipp et al., 2012). Analysis of correlations among power envelopes revealed spatially specific coupling patterns. For instance, signal power was correlated between homologous sensory areas of the two hemispheres (Figure 2D), which matches similar patterns observed in BOLD signals (Figure 1A). Overall, ICMs were most prominent in the alpha and beta band. The power envelope fluctuations were coupled at very slow frequencies below 0.1 Hz (Figure 2C), suggesting a close correspondence to correlated BOLD activity fluctuations (Fox and Raichle, 2007, Deco and Corbetta, 2011 and Raichle, 2010). The data indicate that this approach can reveal a rich set of spectral signatures for functional networks.