EHMCANet

This repo holds code for "Efficient Hierarchical Multiscale Convolutional Attention for Accurate Medical Image Segmentation"

Usage

1. Prepare data

• Download the MoNuSeg dataset from the official website. Then prepare the datasets in the following format for easy use of the code:

  ├── MoNuSeg
    ├── Train_Folder
    │    └──img
    │          ├── TCGA-21-5784-01Z-00-DX1.png
    │          ├── TCGA-21-5786-01Z-00-DX1.png
    │          └── ......
    │    └──labelcol
    │          ├── TCGA-21-5784-01Z-00-DX1.png
    │          ├── TCGA-21-5786-01Z-00-DX1.png
    │          └── ......
    ├── Val_Folder
    │    └──img
    │          ├── TCGA-18-5592-01Z-00-DX1.png
    │          ├── TCGA-AY-A8YK-01A-01-TS1.png
    │          └── ......
    │    └──labelcol
    │          ├── TCGA-18-5592-01Z-00-DX1.png
    │          ├── TCGA-AY-A8YK-01A-01-TS1.png
    │          └── ......
    └── Test_Folder
    │    └──img
    │          ├── TCGA-2Z-A9J9-01A-01-TS1.png
    │          ├── TCGA-44-2665-01B-06-BS6.png
    │          └── ......
    │    └──labelcol
    │          ├── TCGA-2Z-A9J9-01A-01-TS1.png
    │          ├── TCGA-44-2665-01B-06-BS6.png
    │          └── ......

2. Environment

Please prepare an environment with python>=3.7, and then use the command "pip install -r requirements.txt" for the dependencies.

3. Train/Test

• Train

EHMCANet:
python EHMCANet/train.py --root_path ./EHMCANet/train.py --batch_size 4 --img_size 256 

• Test

EHMCANet:
python EHMCANet/test.py --root_path ./EHMCANet/test.py --img_size 256 

You can download the weights of our network from the link.

Reference

Citations