SpindleDetection

SpindleDetection

Paper: SpindleU-Net: An Adaptive U-Net Framework for Sleep Spindle Detection in Single-Channel EEG

Paper link: https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9514837

Content

• Setup

• Dataset Preparation

• Training

Setup

First, clone the repo and install required packages:

git clone https://github.com/LordXX505/SpindleDetection.git
cd SpindleDetection

conda create -n SpindleDetection
conda activate SpindleDetection
pip install -r requirements.txt

Dataset Preparation

We use the Mass SS2 datasets.

You can download from https://doi.org/10.5683/SP3/Y889CS.This is PSG recordings.

And SS2 Sleep Annotations:https://borealisdata.ca/dataset.xhtml?persistentId=doi:10.5683/SP3/Y889CS.

conda activate SpindleDetection
cd SpindleDetection
mkdir ../data/SS2/SS2_bio
mkdir ../data/SS2/SS2_ana
Put PSG recordings to ../data/SS2/SS2_bio and annotations to ../data/SS2/SS2_ana.
sh data.sh
cd utils
python3 augment.py

Training

Using slurm:

sh start.sh

But carefully, you may manually change the Master addr in ./utils/misc the init_distributed_mode function.

And if you want the loss function is CrossEntropy:

sh cross.sh