Dynamic Brain Transformer

Dynamic Brain Transformer is the open-source implementation of the BHI 2023 paper Dynamic Brain Transformer with Multi-level Attention for Functional Brain Network Analysis.

Usage

1. Change the path attribute in file source/conf/dataset/PNC.yaml to the path of your dataset.

2. Run the following command to train the model.

python -m source --multirun datasz=100p model=gfine preprocess=non_mixup project=EGT dataset=ABCD_reg group_attr=w24_s24_mask1_nodo_pos_emb repeat_time=5 model.control=2 model.window_sz=24 model.stride=24 model.mask_0=false model.fc_dropout=false

• datasz, default=(10p, 20p, 30p, 40p, 50p, 60p, 70p, 80p, 90p, 100p). How much data to use for training. The value is a percentage of the total number of samples in the dataset. For example, 10p means 10% of the total number of samples in the training set.

• model, default=(gfine,). Which model to use. The value is a list of model names. For example, gfine means Dynamic Brain Transformer.

• dataset, default=(PNC, ABCD_reg). Which dataset to use. The value is a list of dataset names.

• repeat_time, default=5. How many times to repeat the experiment. The value is an integer. For example, 5 means repeat 5 times.

• preprocess, default=(mixup, non_mixup). Which preprocess to applied. The value is a list of preprocess names. For example, mixup means mixup, non_mixup means the dataset is feeded into models without preprocess.

• model.window_sz, default=360. The window size of the sliding window.

• model.stride, default=360. The stride of the sliding window.

Installation

conda create --name bnt python=3.9
conda install pytorch torchvision torchaudio cudatoolkit=11.3 -c pytorch
conda install -c conda-forge wandb
pip install hydra-core --upgrade
conda install -c conda-forge scikit-learn
conda install -c conda-forge pandas

Citation

Please cite our paper if you find this code useful for your work:

@inproceedings{
  kan2023dbn,
  title={Dynamic Brain Transformer with Multi-level Attention for Functional Brain Network Analysis},
  author={Xuan Kan and Aodong Chen Gu and Hejie Cui and Ying Guo and Carl Yang},
  journal={The IEEE-EMBS International Conference on Biomedical and Health Informatics},
  year={2023},
}