The code of the paper "DOCTer: A Novel EEG-based Diagnosis Framework for Disorders of Consciousness".
DOCTer is a framework designed for diagnosing disorders of consciousness(DOC) primarily using EEG through deep learning methods. For more information about the DOCTer framework, please refer to our paper.
- Python 3.6+
- einops 0.6.1+
- mne 1.1.0+
- numpy 1.20.0+
- pandas 1.4.2+
- pycrostates 0.3.0+
- PyTorch 1.12.0+
- scikit-learn 1.0.2+
- scipy 1.7.3+
Here is a simple example of how to use the DOCTer framework:
epoch=100
datapath="/data/EEG/"
seed=99
fold=10
chs='all'
testf=20
csvfile='./res.csv'
python master_old.py --normalize "y" --chs 'all' --testfreq $testf --csvfile $csvfile --fold $fold --timelen -1 --datapath $datapath --seed 99 --dropout 0.4 --weight_decay 0.0001 --epochs $epoch --batch_size 256 --lr 0.001 --clip 100 --model "DOCTer" --cuda "cuda:0"Thank you for visiting. The dataset cannot be made publicly available upon publication because it contains sensitive personal information. We will continue to improve this repository in the future.
For questions, feedback, or suggestions, please contact us at:
- Email: yue.cao@zju.edu.cn
- GitHub Issues: https://github.com/EEplet/DOCTer/issues
If you find our code is useful, please cite our paper.