skjerns/AutoSleepScorer

An attempt to create a robust sleep scorer using Convolutional Neural Networks with Long Short-Term Memory. This package aims at researchers trying to find an open-source solution for automatic sleep stage classification of human PSG recordings. It is a follow-up of my Master's Thesis: Automatic Sleep Stage Classification using Convolutional Neural Networks with Long Short-Term Memory This package is still under development and not published yet. In this project a Convolutional Neural Network with Long Short-Term Memory is used for the detection of sleep stages. This approach has the advantage that it can automatically detect and extract features from the raw EEG, EMG and EOG signal (see here for example features that are learned by the network).