UC San Diego

Traditional sleep monitoring involves obtaining a polysomnography which uses a minimum of six channels of electrodes to record the following biosignals - electroencephalography, electrooculography, electromyography, electrocardiography and other channels for measuring respiration. This multimodality physiological monitoring poses a lot of discomfort for the person undergoing sleep recording and disrupts the natural sleep owing to the bulk of electrodes and wires used for acquiring the signals, in turn defeating the purpose of monitoring sleep as it naturally occurs. Using a novel, thin, flexible Epidermal Electronic System, unobtrusive sleep monitoring can be performed. Ten healthy adults underwent concurrent sleep EEG recording with epidermal electrodes and conventional electrodes to perform a pilot study, to characterize how similar the clinically relevant aspects of sleep EEG are under the two different recording technologies. Time - frequency spectrogram estimates that can balance temporal and spatial resolution required to understand the EEG acquired from different electrode technologies have been developed. Hypnograms have been generated using state-space probabilistic modeling of sleep EEG from epidermal and conventional electrodes. The outputs of the sleep staging estimators are compared using probabilistic clustering algorithms that operate on the estimates. The outcome of our research demonstrates the capability of epidermal electrodes that can potentially be used as an acquisition cum analysis screening tool for sleep disorders