Hyeong-Jin Kim, Minji Lee, Dong-Ok Won, and Seong-Whan Lee, "CombSleepNet: Combination of Spectral–Temporal CNN and Bi-LSTM with Dense Connectivity for Automatic Sleep Stage Scoring," submit to IEEE Journal of Biomedical and Health Informatics.
- Download the Sleep-EDF database
- Sleep-EDF database is available here.
Note that, we used only 20 people of the Sleep Cassette data.
i.e. from subject SC4001 to SC4192 in the sleep-cassette.
- Sleep-EDF database is available here.
-
Data pre-processing
- Change directory to
./CombSleepNet/pre-processing - Unzip
eeglab.zip - Run
preprocessing.m
- Change directory to
-
Training and testing the CombSleepNet
- Change directory to
./CombSleepNet - Refer to
example.txtto train and test CombSleepNet.
- Change directory to
- Matlab R2019b
- Python3
- Pytorch v1.3.1
- numpy v1.17.4
- scipy v1.3.3
- scikit-learn v0.22
Hypnogram and posterior probability distribution with CombSleepNet for one subject of Sleep-EDF database
Hyeong-Jin Kim
Department of Brain and Cognitive Engineering, Korea University
Email: kme0115@korea.ac.kr
This project is released under the MIT License.