The brain is a very delicate, but sophisticated and complex organ of the body. During life, the brain grows and develops, matures, and then ages like all other organs of the body. This maturation and ageing comes with various physiological and anatomical changes which have an impact on the background activity of the brain. Brain activity can be recorded using various techniques including magnetoencephalography and magnetic resonance imaging, and these recordings combined with signal processing can be useful to characterise the changes in the brain that can be a result of the ageing process. By analysing various aspects of the brain, such as functional and effective connectivity, entropy and complexity, the state of the resting brain at different ages can be understood with greater detail. Furthermore, these analyses can be used to investigate the resting state brain networks that are present in the brain, as well as their topology. This information can be combined with network analysis techniques such as graph theory and used to understand the manner in which the brain both matures and ages.

This research made use of two magnetoencephalogram (MEG) databases recorded from both males and females. The first, containing resting state MEGs (rMEGs) from 220 healthy volunteers (aged 7-84), was used as the main database in this thesis to investigate the effects of age on rMEG signals throughout life. The aim of this research was to make use of rMEG signals and signal processing techniques to determine the effects of healthy ageing on the brain throughout life. It was hypothesised that the effects of age are identifiable using advanced signal processing techniques. Thus, the effects of age on linear interactions, causality, synchronisation, information flow, entropy and complexity, were investigated in the 148 MEG channels lying over 5 brain regions (anterior, central, left lateral, posterior, and, right lateral) using multiple linear and non-linear analysis techniques (namely: Pearson’s correlation, coherence, Granger causality, phase slope index, rho index, transfer entropy, synchronisation likelihood, Lempel-Ziv complexity, permutation Lempel-Ziv complexity, permutation entropy, and, modified permutation entropy). Additionally, graph theory principles were used to evaluate different network components such as integration (global efficiency), segregation (clustering coefficient and modularity), centrality (betweenness), and resilience (strength and assortativity) so as to obtain an understanding of the construct of the resting brain network. Moreover, complex network analysis was also used to determine the overall network topology of the brain network and how this changed at different stages in life. Gender effects were also studied so as to identify if there were any significant differences between males and females at different stages of life. Results from these analyses showed that the healthy resting brain has low effective and functional connectivity, relatively low complexity and entropy, as well as no significant detectable direction of information flow. Therefore this showed that there is very little synchronous or simultaneously occurring information in the rMEG time series. Thus, during rest, the brain resembles a system in limbo/phase transition, with low effective and functional connectivity, relatively low complexity and entropy, and no significant detectable direction of information flow.

The second database used in this research project was obtained during a collaboration visit to the Cognitive and Computational Neuroscience laboratory at the Centre for Biomedical Technology- Universidad Politécnica de Madrid (CTB-UPM). This database was made up of rMEGs recorded from 199 healthy volunteers (aged 60-80), and the focus of this additional set of analyses was to identify differences between the rMEG signals recorded from healthy individuals, those with subjective cognitive decline as well as those with mild cognitive impairment, with the hypothesis that the effects of cognitive decline (namely MCI) are identifiable using advanced signal processing techniques. The objective of this study was to use MEG data and non-linear complexity measures to identify the regions of the brain associated with cognitive decline. Similar to analyses performed using database 1, graph theory and complex network analyses were used to investigate the network structure of the resting brain for controls, and subjects with subjective cognitive decline and mild cognitive impairment. Results from these analyses showed that subjects with mild cognitive impairment had lower complexity scores than those with subjective cognitive decline in broadband analyses. Moreover, in the beta frequency band, controls and subjective cognitive decline subjects had significantly higher MEG complexity than subjects with mild cognitive impairment. A dual effect of hypo- and hyper-connectivity associated with cognitive decline during analyses using Pearson’s correlation and Granger causality was identified, with synchronisation likelihood showing only a decrease associated with mild cognitive impairment. Gender effects were also investigated using this database, and it was found that significant differences were present between males and females with subjective cognitive decline.

All in all, taking the novel findings from this thesis into account, this research project managed to highlight the effects of age, gender and cognitive decline on rMEG signals, and thus an extension to the description of what can be termed an ‘illustration of a physiological rhythm’ showing the evolution of different attributes of the healthy brain activity throughout life was presented. Moreover, in the long term, these results can be used to form a benchmark for healthy ageing which could be used to generate a fingerprint for healthy ageing. This fingerprint can then be used as part of the process of diagnosing pathology, or as a step towards pre-diagnosis of brain pathology.