Simultaneous EEG – functional Near Infrared Spectroscopy (fNIRS) as a new imaging tool for whole night investigation of sleep hemodynamic processes, in young healthy adults

Sleep is a state of reduced mental and physical activity, during which brain activity occurs to support a healthy life. Within each cycle, Rapid Eye Movement (REM) sleep (20% of total sleep time during which most recalled dreams occur), alternates with Non-Rapid Eye Movement (NREM) sleep (80% of total sleep time), subdivided into 3 stages evolving from drowsiness to deep sleep (N1, N2, N3). Measuring synchronized bioelectrical neuronal activity, EEG is the exam of choice to define sleep stages, allowing also the detection of sleep-specific transient oscillations. Bioelectrical neuronal activity is intrinsically associated with cerebral hemodynamic regulations and oxygen metabolism, which can be measured with fMRI or functional Near-Infrared Spectroscopy (fNIRS). Although fMRI combined with EEG has been considered for sleep studies, the strict immobility imposed by MRI limits acquisition duration and the noisy MRI scanner makes it difficult for participants to fall asleep. Alternatively, fNIRS is an emerging and non-invasive wearable brain imaging tool that can be combined with EEG to measure brain activity, offering a unique opportunity for prolonged sleep monitoring. fNIRS measures hemodynamic fluctuations of cortical oxygenated (HbO) and deoxygenated (HbR) hemoglobin concentration changes. Whereas most EEG/fNIRS sleep studies reported in the literature considered only few forehead channels, we are proposing whole night EEG/fNIRS, to assess HbO/HbR cortical fluctuations and memory consolidation processes during healthy sleep, using personalized fNIRS to map hemodynamic processes within the fronto-parietal network. To do so, we applied personalized fNIRS (Cai et al HBM 2021), estimating a personalized locally dense montage, optimizing the position of fNIRS optodes to target bilateral frontoparietal regions, before gluing fNIRS and EEG sensors to the scalp using a clinical adhesive. This approach limits sensitivity to motion and therefore ensures collecting good quality signals during whole-night recordings. We will then promote local 3D reconstruction of HbO/HbR signals within those four regions of interest, by converting scalp measurements into 3D reconstructions along the cortical surface (Cai et al Nature Sc. Report 2022). Therefore, our fNIRS scalp recordings using locally dense montages, could be converted into spatio-temporal neuroimaging results covering bilateral frontal and parietal regions for the whole night recording. For each region, using time-frequency analyses of reconstructed fNIRS signals during the whole night, we are reporting sleep stage specific signatures of HbO/HbR oscillatory patterns, comparing slow hemodynamic oscillatory components in all NREM sleep stages but also in REM phasic and tonic periods, while benefitting from EEG data to assess the contribution of sleep biomarkers (spindles, slow waves) in the generation of those hemodynamic responses. Using this approach, our objective is to build the first atlas of EEG/fNIRS sleep physiology in young healthy adults (18-35 years old). In the second part of the talk, we will present preliminary results from a two-night design that allowed to get insights from memory consolidation processes in the healthy brain, using both bioelectrical (EEG) and hemodynamic (HbO/HbR) signals.