by Prof. Camillo Porcaro (Department of Neuroscience, University of Padova)
When: September 21, 2023, at 3:00 pm
Where: Aula Magna Clinica Neurologica (Via Giustiniani 5, Padova)
In the last two decades, ongoing brain fluctuations at rest (i.e. in the absence of external stimulation and response demands) have consistently been documented to be organised into large-scale networks, each of them characterised by specific structural and functional architectures. These are known as resting-state networks (RSNs) and have been widely reported in numerous neuroimaging and electrophysiological studies. Furthermore, alterations to RSNs have been observed during healthy ageing as well as in many neuropsychiatric and neurological disorders. Specifically, we aimed to investigate how the neuronal dynamics activity at rest can differentiate RSNs and link them to behavioural and perceptual states. Even though linear methods are predominantly used in characterising brain oscillations in healthy and pathological conditions, linear analysis may not be suitable for describing irregular and non-periodic patterns recorded by electrophysiological and neuroimaging techniques. To this end, we characterised the specific neuronal dynamics signature of each RSN, using a complexity measure called Fractal Dimension (FD) that has advantages over classical linear methods such as the well-known fast Fourier transformation (FFT) that are best suited to conditions where the analysed signals are stationary. FD is a general measure of complexity derived from chaos theory, based on the fact that a simple process repeated endlessly becomes very complex, which is the basis for the description of fractals in nature. Knowing that FD is an accurate numerical measure no matter what the properties (stationary, nonstationary, deterministic, or stochastic) of the analysed signal, it is reasonable to accept this advantage over widely used FFT-based or other linear methods.