Kent0n-Li/nnUNet_GUI

可视化 网页版 nnUNet 用户界面 GUI

nnUNet with GUI
可视化网页版本nnUNet

The current code is designed for Windows, not for Linux.

If you already have nnUNet installed, only install the GUI page （如果你已经安装nnUNet，只安装GUI页面）

git clone https://github.com/Kent0n-Li/nnUNet_GUI_tiny.git
cd nnUNet_GUI_tiny
pip install -r requirements.txt

Install from scratch 从零开始安装

Install (安装步骤):

conda create -n nnsam python=3.9
conda activate nnsam

Choose a suitable Pytorch with CUDA to install
根据CUDA选择合适版本的Pytorch进行安装

pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu117
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121

pip install git+https://github.com/ChaoningZhang/MobileSAM.git
pip install timm
pip install git+https://github.com/Kent0n-Li/nnSAM.git

git clone https://github.com/Kent0n-Li/nnUNet_GUI.git
cd Medical-Image-Segmentation-Benchmark
pip install -r requirements.txt

Install from scratch without nnSAM 只安装nnUNet

Install (安装步骤):

conda create -n nnu python=3.9
conda activate nnu

Choose a suitable Pytorch with CUDA to install
根据CUDA选择合适版本的Pytorch进行安装

pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu117
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121

pip install nnunetv2

git clone https://github.com/Kent0n-Li/nnUNet_GUI_tiny.git
cd nnUNet_GUI_tiny
pip install -r requirements.txt

运行可视化软件

python web.py

If you only want to use nnSAM, please install this.
如果你只想运行nnSAM,请访问该代码仓库this

样例数据集：Demo Dataset

Overview 页面总览

Choose Model 选择模型

Import Data 导入你的数据集 （2D: png, 3D: nii.gz)

样例数据集：Demo Dataset

Full Auto 全自动模式，一键完成从数据预处理到训练测试和结果总结

Result Summary 结果总结

Citation

If you find this repository useful for your research, please use the following.

@article{li2023nnsam,
  title={nnSAM: Plug-and-play Segment Anything Model Improves nnUNet Performance},
  author={Li, Yunxiang and Jing, Bowen and Li, Zihan and Wang, Jing and Zhang, You},
  journal={arXiv preprint arXiv:2309.16967},
  year={2023}
}