Univ. Grenoble Alpes, Grenoble, France, CEA Leti, MINATEC Campus, 17 rue des Martyrs, Grenoble, France, Dijon University Hospital, Dpt. Anesth. and Crit. Care, 14 rue Paul Gaffarel, Dijon, France, Grenoble University Hospital, Dpt. Exploration Fonctionnelle du Système Nerveux, Avenue du Maquis du GrésivaudanLa Tronche, France, Grenoble University Hospital, Dpt. Anesth. and Crit. Care, Avenue Maquis du GrésivaudanLa Tronche, France

We present a novel method for automatic sleep scoring based on single-channel EEG. We introduce the use of a deep convolutional neural network (CNN) on raw EEG samples for supervised learning of 5-class sleep stage prediction. The network has 14 layers, takes as input the 30-s epoch to be classified as well as two preceding epochs and one following epoch for temporal context, and requires no signal preprocessing or feature extraction phase. We train and evaluate our system using data from the Sleep Heart Health Study (SHHS), a large multi-center cohort study including expert-rated polysomnographic records. Performance metrics reach the state of the art, with accuracy of 0.87 and Cohen kappa of 0.81. The use of a large cohort with multiple expert raters guarantees good generalization. Finally, we present a method for visualizing class-wise patterns learned by the network. © 2017 Elsevier Ltd

Classification, Convolutional neural network, EEG, Single-channel, Sleep Heart Health Study, Sleep staging

Classification (of information), Convolution, Deep neural networks, Electroencephalography, Neural networks, Sleep research, Automatic sleep scoring, Convolutional neural network, Health-study, Performance metrics, Signal preprocessing, Single channel eeg, Single channels, Sleep staging, Network layers, algorithm, Article, best corrected visual acuity, brain computer interface, cohort analysis, comparative study, controlled study, electroencephalogram, electromyogram, electrooculogram, entropy, eye movement, false negative result, false positive result, human, major clinical study, nerve cell network, predictive value, priority journal, reading, sleep spindle, sleep stage, support vector machine, task performance, theta rhythm