Neuronal phase- and amplitude-coupling in the healthy and diseased human brain

Abstract

Oscillatory neuronal activity has been proposed to facilitate and multiplex the communication between distant brain regions. In particular, two key coupling modes have been discussed in this regard; amplitude- and phase-coupling between distant signals. Both can describe large-scale neuronal interactions independently of one another and have theoretically been linked to distinct neuronal mechanisms. However, to date the direct relations between functional networks derived from these two coupling modes remain unclear. In this thesis we conducted two magnetoencephalography (MEG) studies to empirically assess the relationship between amplitude- and phase-coupling networks in the healthy and diseased human brain. In the first study we analyzed the publicly available human connectome project (HCP S900) dataset of 95 healthy subjects. We applied source-reconstruction to systematically compare cortical amplitude-coupling and phase-coupling patterns in the healthy human brain. We found significant similarities between amplitude- and phase-coupling patterns for almost the entire spectrum and cortex. We further showed that these patterns are similar but non-redundant, indicating a complex spatial and spectral distribution. By combining empirical measurements with simulations and attenuation correction, we sought to ensure that these results were not due to methodological biases but instead reflected relations between genuine amplitude- and phase-coupling, which may indicate at least partially distinct neuronal mechanisms. Additionally, we highlight and clarify the compound nature of amplitude-coupling of orthogonalized signals. In our second study, we measured MEG in 17 relapsing-remitting Multiple Sclerosis patients at an early disease stage (median EDSS = 1.5, range 0 to 3.5) and 17 healthy controls to investigate brain-wide phase- and amplitude-coupling of frequency specific neuronal activity. We developed a new analysis approach that combines dimensionality reduction, bootstrap aggregating and multivariate classification to identify changes of brain-wide coupling in Multiple Sclerosis. We identified systematic and non-redundant changes of both phase- and amplitude-coupling. Changes included both, increased and decreased neuronal coupling in wide-spread, bilateral neuronal networks across a broad range of frequencies. These changes allowed to successfully classify patients and controls with an accuracy of 84%. Furthermore, classification confidence predicted behavioral scores of disease severity. Our results unravel systematic changes of large-scale neuronal coupling in Multiple Sclerosis and suggest non-invasive electrophysiological coupling measures as powerful biomarkers of Multiple Sclerosis. Overall, our two studies provide, to our knowledge, the first systematic analyses describing the relationship between amplitude- and phase-coupling networks. In both, the healthy as well as in the diseased brain, these two coupling modes are related but show distinguishable features. Our findings highlight that amplitude- and phase-coupling might at least partially originate from distinct neuronal mechanisms.