Mandke, Kanad
(2020)
Magnetoencephalographic dynamics and adaptive brain networks across multiple timescales.
PhD thesis, University of Nottingham.
Abstract
The human brain is a complex network of interconnected neurons which function on various scales of time and space. The observable behaviour produced by the human brain is the result of a cascade of several neuronal events that occur over a wide temporal scale. The project investigated processes near the two ends of the temporal scale using magnetoencephalography (MEG), a non-invasive neuroimaging method with high temporal resolution.
As an example of temporal dynamics over the millisecond scale, auditory event-related responses were investigated. To this end, MEG data were recorded as participants performed an auditory oddball task. The data were analysed using Linearly Constrained Minimum Variance beamformer, applied to study non-phase-locked and phase-locked event-related responses. Neuronal sources of sensory-evoked responses, corresponding to phase-locked and non-phase-locked responses, were localised in the bilateral primary auditory cortices and auditory belt area. Whereas, neuronal sources of late, cognitive responses (occurring 300ms after stimulus onset) were localised in modality specific and supra-modal areas, such as the temporo-parietal junction, inferior frontal gyrus, and prefrontal cortex.
As an example of dynamics over years, the project searched for changes in the organisation of brain networks following years of musical training. This research question required the development of a novel method based on graph theory and in particular, the multilayer network framework. This framework treats the human brain as a “network of networks”, allowing us to integrate information from different MEG frequency bands (or imaging modalities) in a unified network description. The central contribution from this thesis is the development of a normalisation method, which allows us to perform multilayer network group comparisons (such as control group vs experimental/clinical group). The efficacy of this method was tested by using simulated brain networks as ground truth.
Subsequently, using the newly developed method, the effects of training related plasticity were investigated in resting-state MEG recordings, without a task requiring the trained skill. Here, we identified that musicians show an integrated network configuration when compared with non-musicians. Furthermore, it was also demonstrated that multilayer networks provide information that more traditional approaches of network analysis cannot, highlighting the benefits of multilayer network analysis in between-group studies.
The thesis concludes with a general discussion on how the multilayer network framework can potentially be used further for studies of cortical dynamics, specifically, multimodal data integration to obtain a holistic picture of the human brain.
Actions (Archive Staff Only)
 |
Edit View |
Tools
Tools
![[img]](/style/images/fileicons/application_pdf.png)