A Reverse Mamba Attention Network for Pathological Liver Segmentation

Overview

Create Environment

conda create -n RMAMamba python==3.8.16
conda activate RMAMamba

Install Dependencies

pip install -r requirements.txt
cd selective_scan && pip install .

Download Checkpoint

Download pretrained checkpoints from Google Drive and move it to the pretrained_pth directory.

Download Dataset

Download CirrMRI600+ dataset from this link and move it to the data directory.

Train

python train.py --model RMAMamba_S

Test

python test.py --model RMAMamba_S

Weight Files

Our weight files and result maps are available on Google Drive.

Citation

Please cite our paper if you find the work useful:

@article{zeng2024rmamamba,
    title={A Reverse Mamba Attention Network for Pathological Liver Segmentation},
    author={Zeng, Jun and Jha, Debesh and Bagci, Ulas},
    bookarticle={xxx},
    year={2024},
}

Contact

Please contact [email protected] for any further questions.