This repo holds the pytorch implementation of DoDNet:
DoDNet: Learning to segment multi-organ and tumors from multiple partially labeled datasets. (https://arxiv.org/pdf/2011.10217.pdf)
Python 3.7
PyTorch==1.4.0
Apex==0.1
batchgenerators
- Clone this repo
git clone https://github.com/jianpengz/DoDNet.git
cd DoDNet
Before starting, MOTS should be re-built from the serveral medical organ and tumor segmentation datasets
| Partial-label task | Data source |
|---|---|
| Liver | data |
| Kidney | data |
| Hepatic Vessel | data |
| Pancreas | data |
| Colon | data |
| Lung | data |
| Spleen | data |
- Download and put these datasets in
dataset/0123456/. - Re-spacing the data by
python re_spacing.py, the re-spaced data will be saved in0123456_spacing_same/.
The folder structure of dataset should be like
dataset/0123456_spacing_same/
├── 0Liver
| └── imagesTr
| ├── liver_0.nii.gz
| ├── liver_1.nii.gz
| ├── ...
| └── labelsTr
| ├── liver_0.nii.gz
| ├── liver_1.nii.gz
| ├── ...
├── 1Kidney
├── ...
Pretrained model is available in checkpoint
- cd `a_DynConv/' and run
CUDA_VISIBLE_DEVICES=0,1 python -m torch.distributed.launch --nproc_per_node=2 --master_port=$RANDOM train.py \
--train_list='list/MOTS/MOTS_train.txt' \
--snapshot_dir='snapshots/dodnet' \
--input_size='64,192,192' \
--batch_size=2 \
--num_gpus=2 \
--num_epochs=1000 \
--start_epoch=0 \
--learning_rate=1e-2 \
--num_classes=2 \
--num_workers=8 \
--weight_std=True \
--random_mirror=True \
--random_scale=True \
--FP16=False
CUDA_VISIBLE_DEVICES=0 python evaluate.py \
--val_list='list/MOTS/MOTS_test.txt' \
--reload_from_checkpoint=True \
--reload_path='snapshots/dodnet/MOTS_DynConv_checkpoint.pth' \
--save_path='outputs/' \
--input_size='64,192,192' \
--batch_size=1 \
--num_gpus=1 \
--num_workers=2
python postp.py --img_folder_path='outputs/dodnet/'
If this code is helpful for your study, please cite:
@inproceedings{zhang2021dodnet,
title={DoDNet: Learning to segment multi-organ and tumors from multiple partially labeled datasets},
author={Zhang, Jianpeng and Xie, Yutong and Xia, Yong and Shen, Chunhua},
booktitle={Proceedings of the IEEE conference on computer vision and pattern recognition},
pages={},
year={2021}
}
Jianpeng Zhang ([email protected])