A Graph-Temporal Fused Dual-Input Convolutional Neural Network for Detecting Sleep Stages from EEG Signals

Sleep is an essential integrant in everyone's daily life. Thereby, it important but challenging problem to construct a reliable and stable system, that can monitor user's sleep quality automatically. In this brief, we combine complex network deep learning propose novel Graph-Temporal fused dual-input Convolutional Neural Network (CNN) method detect stages by using the Sleep-EDF database. Firstly, segment each single-channel EEG signal into non-overlapping 30s epochs set up network. For that, map epoch Limited Penetrable Visibility Graph (LPVG) obtain corresponding Degree Sequence (DS) calculating node degree. Finally, DSs are combined as inputs of CNN learn about graph topology temporal feature representations raw data for purpose classifying two-, three-, four-, five-, six-state. Notably, classification accuracy six-state stage detection 87.21% Kappa value 0.80. The results demonstrate effectiveness our model structure detecting states, whereby they further provide basic strategy future research.