Relationship between large-scale structural and functional brain connectivity in the human lifespan

The relationship between the anatomical structure of the brain and its functional organization is not straightforward and has not been elucidated yet, despite the growing interest this topic has received in the last decade. In particular, a new area of research has been defined in these years, called 'connectomics': this is the study of the different kinds of 'connections' existing among micro- and macro-areas of the brain, from structural connectivity — described by white matter fibre tracts physically linking cortical areas — to functional connectivity — defined as temporal correlation between electrical activity of different brain regions — to effective connectivity—defining causal relationships between functional activity of different brain areas. Cortical areas of the brain physically linked by tracts of white matter fibres are known to exhibit a more coherent functional synchronization than areas which are not anatomically linked, but the absence of physical links between two areas does not imply a similar absence of functional correspondence. Development and ageing, but also structural modifications brought on by malformations or pathology, can modify the relation between structure and function. The aim of my PhD work has been to further investigate the existing relationship between structural and functional connectivity in the human brain at different ages of the human lifespan, in particular in healthy adults and both healthy and pathological neonates and children. These two 'categories' of subjects are very different in terms of the analysis techniques which can be applied for their study, due to the different characteristics of the data obtainable from them: in particular, while healthy adult data can be studied with the most advanced state-of-the-art methods, paediatric and neonatal subjects pose hard constraints to the acquisition methods applicable, and thus to the quality of the data which can be analysed. During this PhD I have studied this relation in healthy adult subjects by comparing structural connectivity from DWI data with functional connectivity from stereo-EEG recordings of epileptic patients implanted with intra-cerebral electrodes. I have then focused on the paediatric age, and in particular on the challenges posed by the paediatric clinical environment to the analysis of structural connectivity. In collaboration with the Neuroradiology Unit of the Giannina Gaslini Hospital in Genova, I have adapted and tested advanced DWI analysis methods for neonatal and paediatric data, which is commonly studied with less effective methods. We applied the same methods to the study of the effects of a specific brain malformation on the structural connectivity in 5 paediatric patients. While diffusion weighted imaging (DWI) is recognised as the best method to compute structural connectivity in the human brain, the most common methods for estimating functional connectivity data — functional MRI (fMRI) and electroencephalography (EEG) — suffer from different limitations: fMRI has good spatial resolution but low temporal resolution, while EEG has a better temporal resolution but the localisation of each signal's originating area is difficult and not always precise. Stereo-EEG (SEEG) combines strong spatial and temporal resolution with a high signal-to-noise ratio and allows to identify the source of each signal with precision, but is not used for studies on healthy subjects because of its invasiveness. Functional connectivity in children can be computed with either fMRI, EEG or SEEG, as in adult subjects. On the other hand, the study of structural connectivity in the paediatric age is met with obstacles posed by the specificity of this data, especially for the application of the advanced DWI analysis techniques commonly used in the adult age. Moreover, the clinical environment introduces even more constraints on the quality of the available data, both in children and adults, further limiting the possibility of applying advanced analysis methods for the investigation of connectivity in the paediatric age. Our results on adult subjects showed a positive correlation between structural and functional connectivity at different granularity levels, from global networks to community structures to single nodes, suggesting a correspondence between structural and functional organization which is maintained at different aggregation levels of brain units. In neonatal and paediatric subjects, we successfully adapted and applied the same advanced DWI analysis method used for the investigation in adults, obtaining white matter reconstructions more precise and anatomically plausible than with methods commonly used in paediatric clinical environments, and we were able to study the effects of a specific type of brain malformation on structural connectivity, explaining the different physical and functional manifestation of this malformation with respect to similar pathologies. This work further elucidates the relationship between structural and functional connectivity in the adult subject, and poses the basis for a corresponding work in the neonatal and paediatric subject in the clinical environment, allowing to study structural connectivity in the healthy and pathological child with clinical data.