Inferring brain network dynamics of MEG and EEG in healthy aging and Alzheimer’s disease

Thesis

Alzheimer’s disease (AD) has long been encumbered by a lack of clinically effective biomarkers, which has impeded its early-stage detection and relegated diagnostic practices to symptom-based observations. Amidst the ongoing quest to discover novel biomarkers with enhanced clinical utility, this thesis aims to elucidate the prospect of static and dynamic changes in brain network features, derived from electroencephalography (EEG) and magnetoencephalography (MEG), to explicate the earliest changes in the brain caused by AD.

The first component of the thesis contrasts MEG and EEG in capturing age-related effects within the healthy aging datasets and validates the methodologies initially designed for MEG for their application to EEG. We demonstrate that while both modalities reveal analogous age effects in static power spectra and narrow-band power, MEG exhibits higher sensitivity to age effects in various brain network features within source space. The adopted analysis techniques successfully replicated previously documented healthy aging effects, substantiating their credibility in inferring transient resting-state network (RSN) dynamics from M/EEG data.

Leveraging the findings above, the second component of the thesis focuses on identifying modality-specific biomarkers for mild cognitive impairment (MCI) and early-stage AD from resting-state M/EEG. The reproducibility of known biomarkers --- evident in static RSN features and the posterior default mode network --- is confirmed in our dataset. Furthermore, we show that the dynamic effects of MCI and early-stage AD in the wide-band power and temporal dynamics of RSNs can be recognized as potential AD biomarkers. These results are among the first to demonstrate the statistically significant effects of prodromal AD in resting-state M/EEG.

Together, our findings provide a concrete case that M/EEG-derived static and dynamic RSN features can potentially be reliable and clinically meaningful biomarkers. This research propels the field towards refining early detection mechanisms for AD, laying a foundation for ensuing research to build upon, and steering future diagnostic practices towards enhanced efficacy.

Authors: Cho, S

• DOI: 10.5287/ora-44k06b6ad
• Type of award: MSc by Research
• Level of award: Masters
• Awarding institution: University of Oxford
• Language: English
• Keywords: electroencephalography; aging; magnetoencephalography; Alzheimer's disease; brain network dynamics
• Subjects: Computational neuroscience; Neuropsychiatry; Cognitive neuroscience