| Summary: | Automatic and real-time sleep scoring is necessary to develop user interfaces that trigger stimuli in specific sleep stages. However, most automatic sleep scoring systems have been focused on offline data analysis. We present the first, real-time sleep staging system that uses deep learning without the need for servers in a smartphone application for a wearable EEG. We employ real-time adaptation of a single channel Electroencephalography (EEG) to infer from a Time-Distributed Convolutional Neural Network (CNN). Polysomnography (PSG) -the gold standard for sleep staging-requires a human scorer and is both complex and resource-intensive. Our work demonstrates an end-to-end, smartphone-based pipeline that can infer sleep stages in just single 30-second epochs, with an overall accuracy of 83.5% on 20-fold cross validation for 5-stage classification of sleep stages using the open Sleep-EDF dataset. For comparison, inter-rater reliability among sleep-scoring experts is about 80% (Cohen's k=0\pmb.68 to \pmb0.76). We further propose an on-device metric independent of the deep learning model which increases the average accuracy of classifying deep-sleep (N3) to more than 97.2% on 4 test nights using power spectral analysis. Keyword: Sleep; Electroencephalography; Real-time systems; Brain modeling; Electrooculography; Spectral analysis; Training
|
|---|
