Fluctuations in physiological and neurological data within and between states of consciousness: The relevance for the management of patients with a disorder of consciousness

The prolonged nature of a disorder of consciousness (DoC) instills the view that the behavior and governing brain dynamics have a static nature. However, previous research has shown that the clinical diagnosis is not stable (Wannez et al., 2017). At least five different assessments were needed to achieve 95\% accuracy in the diagnosis, meaning that patients likely fluctuate between different states of consciousness. By extension, there should be dynamics within a state that allow for these potential changes. The main aim of this thesis is to better understand these fluctuations. The review in the first chapter shows that sleep, the main natural fluctuation in consciousness, is difficult to define, yet severely affected and diagnostically relevant. When empirically investigating the sleep-wake cycle through actigraphy, the second chapter shows that circadian rhythms are often replaced by faster ultradian rhythms in DoC. Importantly, tracking the remaining rhythms could potentially be used to reduce the behavioral misdiagnosis. At the same time, the tracked rhythms answered a crucial question by showing the robustness of Positron Emission Tomography (PET) to these behavioral fluctuations. With the altered sleep-wake cycle firmly established, the third chapter puts its consequences into perspective by demonstrating a potential link between insomnia and DoC. More specifically, the consequence of both are remarkably similar and might thus bias individual investigations. Next, the thesis turns from between-state sleep-wake cycle fluctuations to within-state fluctuations in chapter four. It summarizes the state-of-the-art in brain state identification, showing slower, less complex and less efficient functional communication in unconsciousness, and puts forward a framework to improve its integration with neuromodulation in order to promote consciousness. In the remaining chapters, the within-state fluctuations of the brain state are investigated on multiple levels, as recommended in the previous review. The fifth chapter characterizes a driving force of fluctuations, endogenous events in the electroencephalography (EEG), showing increased integration compared to non-events. Moreover, it highlights that in consciousness the effect of an event is both temporally (i.e., faster) and spatially (i.e., more consistent) efficient. In a similar vein, over groups of these events, the brain seemed to operate efficiently, closer to a critical state. The sixth and final chapter reconciled the deeper fundamental understanding of fluctuations with the need for clinical relevance by identifying dynamic EEG regimes that can be visually scored based on the EEG power spectrum. The visual evaluation proved reliable, transferable through developed scoring guidelines and validated by objective measurement of the power spectrum. Importantly, putative better regimes are associated with increased consciousness. A group of patients who are at the bedside unconscious but have a potential for covert consciousness, based on separate PET assessment, also performed on par with their conscious-at-the-bedside counterpart. Finally, glucose metabolism as measured through PET was also associated with both the dynamic regimes and with regional, source reconstructed relative EEG power.