/DuMeta

Code for our DuMeta

Primary LanguagePythonApache License 2.0Apache-2.0

Dual Meta-Learning with Longitudinally Generalized Regularization for One-Shot Brain Tissue Segmentation Across the Human Lifespan

by Yongheng Sun, Fan Wang, Jun Shu, Haifeng Wang, Li Wang, Deyu Meng, Chunfeng Lian.

Introduction

This repository is for our ICCV 2023 paper 'Dual Meta-Learning with Longitudinally Generalized Regularization for One-Shot Brain Tissue Segmentation Across the Human Lifespan'.

A. You can follow the steps below to pretrain and finetune your model on your data with DuMeta

1 Preparation

1.1 Install nnUNetv1.

1.2 Process your data as required by nnUNetv1.

nnUNet_plan_and_preprocess -t XXX --verify_dataset_integrity

1.3 Run nnUNetv1 baseline on your data. Save your Generic_UNet, DataLoader3D, and get_moreDA_augmentation hyperparameters.

nnUNet_train 3d_fullres nnUNetTrainerV2 TASK_NAME_OR_ID FOLD (additional options)

1.4 Replace your nnUNet folder with this repository (I rewrote the files in the folders of \nnunet\network_architecture and \nnunet\training\network_training).

1.5 Import Generic_UNet, DataLoader3D, and get_moreDA_augmentation with hyperparameters in step 1.3, as in main.py.

2 Training and Testing

2.1 Meta-train (Pretrain).

python main.py

2.2 Meta-test (Finetune).

Load model parameters of step 2.1, and run baseline as step 1.3.

B. Or you can load our pretrained model and finutune on your own data.

  1. You can download the pretrained model via https://drive.google.com/file/d/1t6nCM376LBVHXjktr52k8KeDwuZZTLy2/view?usp=drive_link.

  2. Process your data as required by nnUNetv1 using below command:\

    nnUNet_plan_and_preprocess -t XXX --verify_dataset_integrity

  3. then change nnUNetPlansv2.1_plans_3D.pkl to change patch size of input:\

    python change_plans.py

  4. finetune your model using below command:\

    nnUNet_train 3d_fullres nnUNetTrainerV2_FT TASK_NAME_OR_ID FOLD (additional options)

Acknowledgement

The code is based on nnUNetv1 (https://github.com/MIC-DKFZ/nnUNet/tree/nnunetv1).

Citation

If you find this project useful for your research, please consider citing:

@inproceedings{sun2023dual,
  title={Dual Meta-Learning with Longitudinally Consistent Regularization for One-Shot Brain Tissue Segmentation Across the Human Lifespan},
  author={Sun, Yongheng and Wang, Fan and Shu, Jun and Wang, Haifeng and Wang, Li and Meng, Deyu and Lian, Chunfeng},
  booktitle={Proceedings of the IEEE/CVF International Conference on Computer Vision},
  pages={21118--21128},
  year={2023}
}