Thanks for reaching out! Here's the explanation for ECG-Mamba and Non-Uniform-Mix.
Quick note: This Markdown file is currently being updated and should be finished very soon. Please check back shortly!
Introduction
Citing
Please cite our paper(s) if you find this repository useful.
@article{jiang2025ecg,
title={ECG-Mamba: Cardiac Abnormality Classification with Non-Uniform-Mix Augmentation on 12-Lead ECGs},
author={Jiang, Huawei and Mutahira, Husna and Wei, Shibo and Muhammad, Mannan Saeed},
journal={IEEE Journal of Translational Engineering in Health and Medicine},
year={2025},
publisher={IEEE}
}
If you are interested in this area, you can also cite the below paper on 12-lead ECG multi-label classification via Mamba architecture. This work demonstrated an increase of 3% in AUPRC compared to the ECG-Mamba model.
@misc{jiang2025dimensionalcnnecgmamba,
title={One Dimensional CNN ECG Mamba for Multilabel Abnormality Classification in 12 Lead ECG},
author={Huawei Jiang and Husna Mutahira and Gan Huang and Mannan Saeed Muhammad},
year={2025},
eprint={2510.13046},
archivePrefix={arXiv},
primaryClass={cs.CV},
url={https://arxiv.org/abs/2510.13046},
}
Getting started
Dataset
In this paper, two challenge datasets are used: the PhysioNet/CinC Challenges of 2020 and 2021. The datasets can be downloaded from the website below:Click
Required package
ECG-Mamba, adapted from Vision Mamba (ViM), requires CUDA 11.8 for compatibility. Follow this link: https://github.com/hustvl/Vim/issues/53 to install the necessary packages
How to run
you can go to the folder of scripts to run the program.
for the scenario challenge 2020
bash ECG_scenario2020.sh
for the scenario challenge 2021
bash ECG_scenario2021.sh
Contact
If you have a question, please start a discussion in the Community section (this is similar to opening an issue on GitHub).
I will do my best to help you out, just as I received so much support at the beginning of my own research journey.
