SNNUBIAI / SCAAE

Discovering Dynamic Functional Brain Networks via Spatial and Channel-wise Attention. Submitted to 36th Conference on Neural Information Processing Systems (NeurIPS 2022)

SCAAE

• Discovering Dynamic Functional Brain Networks via Spatial and Channel-wise Attention arxiv:2205.09576

Download Training Data

A simple way to download part of ADHD-200 datasets is to use nilearn. These data can be used as training data.

from nilearn import datasets
adhd_dataset = datasets.fetch_adhd(n_subjects=40, "./data/")

You can make the dataset by:

from dataset import LoadADHD200
data = LoadADHD200(img_path="./data/adhd/data/", 
                    mask_path="./data/ADHD200_mask_152_4mm.nii.gz",
                    save_fmri=True, 
                    save_path="./data/adhd200.npy")

More preprocessed ADHD-200 data can be accessed here ADHD-200 Preprocessed.

Training

• Training model

chmod +x train.sh
./train.sh

• Tensorboard

tensorboard --logdir=./logdir/

Result

The result in ADHD-200 dataset shown in constructFBN.ipynb

The result in task-based fMRI dataset shown in task-gambling-avg-group-wise-40.ipynb.

We use the Baseline developed by our SNNUBIAI Lab for evaluation.

Citing SCAAE

@article{liu2022discovering,
  title={Discovering Dynamic Functional Brain Networks via Spatial and Channel-wise Attention},
  author={Liu, Yiheng and Ge, Enjie and He, Mengshen and Liu, Zhengliang and Zhao, Shijie and Hu, Xintao and Zhu, Dajiang and Liu, Tianming and Ge, Bao},
  journal={arXiv preprint arXiv:2205.09576},
  year={2022}
}

Related Work

• STCA

@inproceedings{liu2023spatial,
  title={Spatial-Temporal Convolutional Attention for Mapping Functional Brain Networks},
  author={Liu, Yiheng and Ge, Enjie and Qiang, Ning and Liu, Tianming and Ge, Bao},
  booktitle={2023 IEEE 20th International Symposium on Biomedical Imaging (ISBI)},
  pages={1--4},
  year={2023},
  organization={IEEE}
}

@article{liu2024spatial,
  title={Spatial-temporal convolutional attention for discovering and characterizing functional brain networks in task fMRI},
  author={Liu, Yiheng and Ge, Enjie and Kang, Zili and Qiang, Ning and Liu, Tianming and Ge, Bao},
  journal={NeuroImage},
  pages={120519},
  year={2024},
  publisher={Elsevier}
}