Abstract

The proposed research will focus on the classification of the brain states of sleep, sleep stages and wakefulness as well as on the brain state-dependent classification of activated memory representations, based on combined recordings of fMRI and EEG signals, an area of research that is not well developed in Chile. The research originates from evidence indicating that reactivation of memory representations during slow wave sleep (SWS) consolidate hippocampus-dependent declarative memories, i.e., facilitates their storage into long-term. The present research aims at decoding patterns of reactivated memory representations that are assumed to depend on the brain state. For this purpose, in a first step pattern classification techniques for decoding the brain states of sleep shall be developed. In a second step, patterns associated with the (re-) activation of emotional and non-emotional declarative and of procedural memories during wakefulness and in the different sleep stages shall be identified. Memories will be reactivated using odor cuing. A combination of EEG and fMRI signals will be used for decoding to increase accuracy of pattern identification. Pattern classification will be performed in real time to enable immediate manipulations of activated memory representations, thereby validating the approach. This work will not only further our understanding of memory processing in sleep, but also could potentially be applied for innovative treatments in sleep and memory disorders by brain state detection and manipulation. The combined expertise of the PI and collaborators in simultaneous multimodal neuroimaging, pattern classification of brain states of these signals for neuroscience research and Brain-Computer Interface applications, and a strong track-record of sleep and memory research will serve to fulfill the objectives of this project. Furthermore, recent developments and publications in the PI’s lab in Chile in relevant areas make simultaneous fMRI/EEG measurements in sleep feasible and compelling.