U-MLP

Pytorch Code base for “U-MLP: MLP-based Ultralight Refinement Network for Multimodal Medical Image Segmentation”,The full code will be published after the article is accepted! Thank you for your attention

Introduction

The convolutional neural network (CNN) and Transformer play an important role in computer-aided diagnosis and intelligent medicine. However, CNN cannot obtain long-range dependence, and Transformer has the shortcomings in computational complexity and a large number of parameters. Recently, compared with CNN and Transformer, the Multi-Layer Perceptron (MLP)-based medical image processing network can achieve higher accuracy with smaller computational and parametric quantities. Hence, in this work, we propose an encoder-decoder network, U-MLP, based on the ReMLP block.

Using the code:

The code is stable while using Python 3.8.0, CUDA >=11.1

Clone this repository:

git clone https://github.com/xielaobanyy/U-MLP
cd U-MLP

To install all the dependencies :

conda env create U-MLP python==3.8.0
conda activate U-MLP
pip install -r requirements.txt

Datasets

ISIC-2018 - Link
MSD-Spleen - Link
MSD-heart - Link
CHAOS - Link

Data Format

Make sure to put the files as the following structure (e.g. the number of classes is 2):

inputs
└── <dataset name>
    ├── images
    |   ├── 001.png
    │   ├── 002.png
    │   ├── 003.png
    │   ├── ...
    |
    └── masks
        ├── 0
        |   ├── 001.png
        |   ├── 002.png
        |   ├── 003.png
        |   ├── ...

For binary segmentation problems, just use folder 0.

Training and Validation

Train the model.

python train.py --dataset <dataset name> --arch resmlp_12

Evaluate.

python val.py --name <exp name>

Acknowledgements：

This code-base uses certain code-blocks and helper functions from UNeXt Link.

Citation:

@article{Shuo Gao U-MLP,
  title={U-MLP: MLP-based Ultralight Refinement Network for Multimodal Medical Image Segmentation},
  author={Shuo Gao, Wenhui Yang, Menglei Xu, Hao Zhang, Hong Yu, Airong Qian, Wenjuan Zhang},
  journal={Computers in Biology and Medicine},
  DOI={https://doi.org/10.1016/j.compbiomed.2023.107460}
}

xielaobanyy / U-MLP