Sweet Dreams Are Made of These: Investigating the Neural Mechanisms of Dream Traits and States

Abstract

Dreaming is a universal yet poorly understood phenomenon. This thesis investigates the neural mechanisms of dream traits and states using multimodal neuroimaging, including structural and functional MRI, diffusion imaging, and simultaneous EEG/fMRI. By addressing methodological challenges and leveraging large-scale datasets, the studies provide novel insights into the neural correlates of dream recall and dreaming during sleep. A systematic review highlights key obstacles in sleep neuroimaging, such as small sample sizes and the difficulty of capturing REM sleep in the MRI environment. Empirical work demonstrates the absence of robust associations between nightmare frequency and brain connectivity, underscoring the need for rigorous methodology and replication. Using a multimodal fusion approach, findings show that frequent dream recall is linked to enhanced white matter integrity in memory and sensory integration regions, alongside reduced gray matter volume in primary visual areas. Furthermore, state-dependent analyses reveal that dreaming relies on dynamic interactions between the default mode, frontoparietal, and ventral attention networks during NREM sleep, rather than limbic or subcortical processes alone. Together, these results support a continuum model of internally generated cognition, advancing our understanding of the neural basis of dreaming and its relevance to sleep, memory, and consciousness research.