The modeling of phase ICMs has just begun (David and Friston, 2003 and Battaglia et al., 2012), and a systematic theoretical analysis of these spectral

coupling modes and their interaction with envelope ICMs still presents a challenge. Another challenge for modeling is to describe the impact of network history on ICMs. Pilot models have demonstrated that mechanisms such as spike-timing-dependent plasticity may contribute to shaping ICMs. For example, in a model of spiking neurons, Izhikevich et al. (2004) found that the interplay between spike-timing-dependent plasticity and conduction learn more delays led to the formation of modules of strongly connected neurons capable of producing time-locked spikes. Alternatively, modular connectivity could be produced from a combination of synchronization-dependent plasticity and growth-dependent plasticity in a neural mass model (Stam et al., 2010). More detailed models will be required to show precisely how previous functional synchronization becomes encoded in patterns of structural connectivity and corresponding ICMs.

A key goal for future modeling approaches will also be to explain the alterations of ICMs in neuropsychiatric disorders. As discussed in the preceding section, even focal stroke typically has a spatially widespread impact on network dynamics and ICMs. This can be modeled by considering the effect of focal lesions of nodes and their connections on envelope ICMs (Alstott et al., 2009). A recent study investigating the impact selleck products of moderate, but spatially

unspecific, disconnection has demonstrated a decrease in small-world properties and global integration reminiscent of the changes observed in schizophrenia (Cabral et al., 2012). Computational approaches may also become relevant for understanding alterations of ICMs in Liothyronine Sodium other network diseases, such as MS. Several computational models suggest that a shift of conduction delays away from the normal set point may lead to suboptimal exploration of the dynamical attractor landscape (Ghosh et al., 2008). The studies reviewed in the preceding sections comply with the notion that the brain’s dynamics are to a large extent determined by its intrinsic communication but much less by interactions with its environment. They demonstrate that intrinsic coupling modes are present in ongoing activity that reflects the sedimented results of previous learning, encodes relevant priors for future processing, and predicts perception and behavior both in the healthy organism and in disorders that affect brain networks. The available data support a differentiation between two types of ICMs (Table 1) that seem to reflect the operation of distinct coupling mechanisms and have therefore been termed “envelope ICMs” and “phase ICMs.” While the latter arise from phase coupling of band-limited oscillatory signals, the former are best described as coupled aperiodic fluctuations of signal envelopes.