Analysis of multi-frequency dynamic coherence networks in large-scale electrophysiological recordings
Tremendous technological progress in the capacity to record simultaneously from multiple brain regions is making increasingly necessary to adopt a system’s perspective when describing brain function. Distant brain regions interact between them in a dynamic manner even to perform very simple tasks and flexibly changing patterns of interactions can be described as time-changing multiplex directed networks, where different layers describe transient inter-regional oscillatory coherence in multiple frequency bands (or across bands). In the framework of this project we propose to study various datasets (from rodent to non-human primate) gathered during large-scale electrophysiological experiments and, more specifically, to analyse them adopting sophisticated tools from complex network analyses that, initially forged in statistical physics or the science of social networks, have not yet been adapted to be compliant with neural datasets.
We will characterize how spatio-tempo-spectral network patterns correlate or predict behaviour and cognitive performance (e.g. in working memory or perceptual tasks) or how they are altered in specific pathological models (e.g. ALS). Network analyses of multi-scale electrophysiological data will be complemented by the design of connectome-based computational models whose simulated oscillatory coherence networks will be compared with the empirically measured one, to reverse engineer possible physiological mechanisms (e.g. alterations of excitability) that may underpin generalized functional connectivity changes.