Investigation of physiological and pathological conditions using electroencephalographic connectivity metrics

Functional connectivity (FC) metrics identify statistical (undirected)associations among distinct brain areas and thereforerepresent a powerful tool to investigate brain inter-regionalinteractions in distinct behavioural states. However, the applicationand interpretation of FC in electrophysiological datais impacted by important confounds related to the instantaneouspropagation of electric fields generated by primarycurrent sources to many of the on-scalp sensors – the so-calledphenomenon of “volume conduction”. Because of this linearmixing of different sources, common FC methods maylead to the identification of apparent couplings that do not reflecttrue brain inter-regional interactions. To overcome thisproblem, new FC metrics have been specifically designed tominimize the impact of volume conduction. Among thesenovel methods, the weighted Phase Lag Index (wPLI) andthe weighted Symbolic Mutual Information (wSMI) attracteda growing interest during the last decade, and have been successfullyapplied to describe brain function in a wide rangeof different conditions, including states associated with alteredlevels of consciousness. In spite of the many promisingapplications and results, the two methods have never beencharacterized in detail, nor compared to investigate their potentialsimilarities or differences. Given these premises, inthe present thesis, my aim was to assess the properties ofwPLI and wSMI in order to define their respective potentialadvantages and disadvantages, as well as to determinewhether useful information could be gained through theircombined application. To this aim I performed three distinct,complementary studies. In my first project, I simulatedrealistic high-density EEG data based on imposed interactiondynamics between sources of interest to test the accuracyof wPLI and wSMI at detecting different types of linearand nonlinear functional interactions. Based on the resultingfinding that they provide complementary information,I applied the two methods to the study of EEG data, collectedin physiological and pathological states. In my secondstudy, I analyzed power, wPLI and wSMI changes across distinctphysiological stages of vigilance, specifically wakefulness(W), NREM and REM sleep in 24 healthy participants.Specifically, I explored the role of power- and FC-based featuresin identifying differences between all stages of interest(W, N2, N3, REM), stages characterized by higher (W+REM)and lower (N2+N3) probabilities of conscious experiences anddifferences in sensory disconnection (REM vs. W), using across-participant classification paradigm. Finally, in my thirdstudy, I applied the two methods for investigating the effectsof motor rehabilitation on brain functional correlates in 16multiple sclerosis patients. Obtained results demonstratedthat wPLI and wSMI provide distinct and complementary informationabout functional brain dynamics and indicate thatthe conjoint use of these two methods may represent a powerfultool to investigate brain connectivity in physiological andpathological conditions.