Configurable Mobile System for Autonomous High-Quality Sleep Monitoring and Closed-Loop Acoustic Stimulation

Abstract

Present-day sleep research in humans is largely dependent on complex and costly laboratory setups, which require controlled supervision. As it is highly desirable to study sleep and to monitor sleep interventions in a realistic setting at home, new mobile approaches with equivalent performance to lab-based systems are needed. We present here the development and evaluation of a mobile system for sleep-biosignal monitoring and real-time intervention for ambulatory sleep research. We evaluated the system for electroencephalogram (EEG) signal quality and compared it to an established sleep EEG recording system. The real-time EEG signal processing performance was evaluated by implementing a closed-loop auditory deep-sleep stimulation algorithm, and we calculated the precision of slow wave (SW) phase targeting during 93 nights. The obtained EEG signals contained similar power spectrograms and high correlations in the delta (0.98) and sigma (0.99) bands when compared to the reference system. The SW phase targeting [mean 44.6°, standard deviation (SD) 46.8°] was comparable to previously published, lab-based approaches. We have, thus, demonstrated that our device is suitable for performing unobtrusive, multinight monitoring and intervention at home.