The human brain can be divided into multiple areas, each responsible for different aspects of behaviour. For a century we have been developing techniques to non-invasively map these areas and their associated functions, a discipline now known as neuroimaging. In recent years the field has undergone a paradigm shift to investigate how the brain communicates with itself; it is widely regarded that healthy brain function relies upon efficient connectivity between different functional areas, and the neuroimaging field has been revolutionised by our ability to estimate this connectivity. Studies into communication between spatially separate locations in the brain have revealed a series of robust functional networks which govern mental processes. However these studies have been based on the temporal averaging of minutes or even hours of data to give us a generalised ’snapshot’ of connectivity. Increasing evidence shows us that these connections are dynamic in space, time and frequency and so the next generation of of neuroimaging methods, which capture this 5-dimensional connectivity will prove to be key tools in the investigation of brain networks and ultimately their breakdown in disease. In this thesis we introduce novel methods to capture non-stationarity using magnetoencephalography (MEG), an imaging modality which measures the changes in extracranial magnetic fields associated with neuronal current flow. MEG is a direct measurement of neural activity and has an excellent temporal resolution, which makes it attractive for non-invasively tracking dynamic functional connections. However there are many technical limitations which can confound assessment of functional connectivity which have to be addressed. In Chapters 2 and 3 we introduce the theory behind MEG; specifically how it is possible to measure the femtoTelsa changes in magnetic field generated by the brain and how to project these data to generate a 3-dimensional picture of current in the brain. Chapter 4 reviews some of popular methods of assessing functional connectivity and how to control for the influence of artefactual functional connections erroneously produced during source projection. Chapter 5 introduces a pipeline to assess functional connections across time, space and frequency and in Chapter 6 we apply this pipeline to show that resting state networks, measured using ’static’ metrics are in fact comprised of a series of rapidly forming and dissolving subnetwork connections. Finally, Chapter 7 introduces a pipeline to track dynamic network behaviour simultaneously across the entire brain volume and shows that networks can be characterised by their temporal signatures of connectivity.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:692739 |
| Date | January 2016 |
| Creators | O'Neill, George C. |
| Publisher | University of Nottingham |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://eprints.nottingham.ac.uk/33502/ |
Page generated in 0.0023 seconds