Sleep as a Window into Multiscale Brain Coordination in Physiological and Pathological Conditions

Objective. This thesis investigates how large-scale brain activity is coordinated across physiological brain states and how this coordination is reshaped by neurological disorders. Sleep and wakefulness provide a structured physiological framework in which patterns of neural interaction emerge across spatial, temporal, and frequency scales. However, it remains unclear whether brain pathology disrupts neural coordination in an unstructured manner or reorganizes it within the intrinsic state-dependent architecture of brain dynamics. To address this question, this work examines two distinct neurological conditions—focal ischemic stroke and epilepsy—which alter brain networks through different mechanisms, but both affect large-scale neural coordination. Approach. To investigate these processes, I combined longitudinal intracortical electrophysiological recordings in a rodent model of focal ischemic stroke with intracranial stereoelectroencephalography (SEEG) recordings from epilepsy patients. This framework enabled the characterization of neural coordination across multiple temporal, spatial, and frequency scales, both in rodent and human brain. Brain synchronization was quantified using phase synchronization (PLV) and phase–amplitude coupling (PAC), capturing interactions both within and across frequency bands. In the rodent model, analyses focused on slow-wave sleep (SWS), enabling the investigation of post-lesional network reorganization within a stable physiological regime. In humans, analyses were conducted across vigilance states (wakefulness, NREM, and REM sleep), allowing the characterization of how physiological and pathological dynamics unfold within different brain states. Main results. The results demonstrate that brain coordination is modulated by vigilance states, which define distinct regimes of large-scale interaction across cortical regions and frequency bands. In the rodent model of ischemic stroke, structural perturbation did not lead to a simple loss of connectivity but induced a transient increase in low-frequency synchronization and phase amplitude coupling during early recovery, followed by partial normalization at later stages, indicating a structured reorganization of network dynamics. In humans, physiological coordination patterns differed markedly across wakefulness and sleep, with NREM sleep characterized by enhanced low-frequency synchronization and hierarchical cross-frequency interactions, and REM sleep showing reduced large-scale coupling. In epilepsy, pathological activity was associated with increased synchronization and altered phase amplitude coupling, particularly in slow and high-frequency bands. However, these alterations remained embedded within the same state-dependent coordination framework observed in physiological conditions and were modulated by vigilance states, with REM sleep showing a reduced expression of pathological coupling. Significance. Overall, these findings demonstrate that large-scale brain coordination is intrinsically shaped by vigilance states across both physiological and pathological conditions. Rather than disrupting neural dynamics in an unstructured manner, stroke- and epilepsy-related alterations unfold within the same state-dependent organizational framework that characterizes physiological brain activity.By establishing sleep as a privileged physiological context for investigating neural coordination, this work provides a unifying perspective on how brain networks reorganize across health and disease and highlights the importance of considering brain state when interpreting pathological activity and large-scale neural interactions.