Functional brain networks: intra and inter-subject variability in healthy individuals and patients with neurological or neuropsychiatric diseases.

The projects of this thesis sits at the intersection between classical neuroscience and aspects related to engineering, signals’ and neuroimaging processing. Each of the three years has been dedicated to specific projects carried out on distinct datasets, groups of individuals/patients and methods, putting great emphasis on multidisciplinarity and international mobility. The studies carried out in Cagliari were based on EEG (electroencephalography), and the one conducted abroad was developed on functional magnetic resonance imaging (fMRI) data. The common thread of the project concerns variability and stability of individuals' features related primarily to functional connectivity and network, as well as to the periodic and aperiodic components of EEG power spectra, and their possible use for clinical purposes. In the first study (Fraschini et al., 2019) we aimed to investigate the impact of some of the most commonly used metrics to estimate functional connectivity on the ability to unveil personal distinctive patterns of inter-channel interaction. In the second study (Demuru et al., 2020) we performed a comparison between power spectral density and some widely used nodal network metrics, both at scalp and source level, with the aim of evaluating their possible association. The first first-authored study (Pani et al., 2020)was dedicated to investigate how the variability due to subject, session and task affects electroencephalogram(EEG) power, connectivity and network features estimated using source-reconstructed EEG time-series of healthy subjects. In the study carried out with the supervision of Prof. Fornito (https://doi.org/10.1016/j.pscychresns.2020.111202) during the experience at the Brain, Mind and Society Research Hub of Monash University, partial least square analysis has been applied on fMRI data of an healthy cohort to evaluate how different specific aspects of psychosis-like experiences related to functional connectivity. Due to the pandemic of Sars-Cov-2 it was impossible to continue recording the patients affected by neurological diseases (Parkinson’s, Diskynesia) involved in the study we planned for the third year, that should have replicated the design of the first first-authored one, with the aim of investigate how individual variability/stability of functional brain networks is affected by diseases. For the aforementioned reason, we carried out the last study on a dataset we finished to record in February 2020. The analysis has the aim of investigate whether it is possible by using 19 channels sleep scalp EEG to highlight differences in the brain of patients affected by non-rem parasomnias and sleep-related hypermotor epilepsy, when considering the periodic and aperiodic component of EEG power spectra.