DM6190 Assignment 1: Vestibular Schwannoma and Cochlea Segmentation from Contrast-enhanced T1 MRI

Team Member

• Zhao Ziyuan
• Ng Han Wei

Abstract

Segmentation is one of the most important steps in medical image analysis. It aims to accurately divide the image into meaningful groups. Recently, deep convolutional neural networks, particularly fully convolution networks, have achieved state-of-the-art results on semantic image segmentation. In this work, we mainly reviewed two representative fully convolution networks, i.e., FCN and UNet, and implemented them in one challenging segmentation task, vestibular schwannoma and cochlea segmentation.

Reimplemented methods

• FCN, UNet, and DRN

Setup

1. Follow official guidance to install Pytorch.
2. Clone the repo
3. cd code

Data Preparation

Cross-Modality Domain Adaptation for Medical Image Segmentation Challenge (CrossMoDA) dataset https://crossmoda-challenge.ml/

Only source training dataset (contrast-enhanced T1) was used.

Please use utils/crossmoda/preprocess.ipynb to explore the data preprocessing process.

Training

Run bash train.sh

Evaluation

Run validation.py

Visualization

Please use utils/visualization.ipynb

Citation

If you find the codebase useful for your research, please cite the papers:

@inproceedings{long2015fully,
  title={Fully convolutional networks for semantic segmentation},
  author={Long, Jonathan and Shelhamer, Evan and Darrell, Trevor},
  booktitle={Proceedings of the IEEE conference on computer vision and pattern recognition},
  pages={3431--3440},
  year={2015}
}

@inproceedings{ronneberger2015u,
  title={U-net: Convolutional networks for biomedical image segmentation},
  author={Ronneberger, Olaf and Fischer, Philipp and Brox, Thomas},
  booktitle={International Conference on Medical image computing and computer-assisted intervention},
  pages={234--241},
  year={2015},
  organization={Springer}
}

@inproceedings{yu2017dilated,
  title={Dilated residual networks},
  author={Yu, Fisher and Koltun, Vladlen and Funkhouser, Thomas},
  booktitle={Proceedings of the IEEE conference on computer vision and pattern recognition},
  pages={472--480},
  year={2017}
}

@inproceedings{zhao2021mt,
  title={MT-UDA: Towards Unsupervised Cross-modality Medical Image Segmentation with Limited Source Labels},
  author={Zhao, Ziyuan and Xu, Kaixin and Li, Shumeng and Zeng, Zeng and Guan, Cuntai},
  booktitle={International Conference on Medical Image Computing and Computer-Assisted Intervention},
  pages={293--303},
  year={2021},
  organization={Springer}
}

Acknowledgement

Part of the code is adapted from open-source codebase and original implementations of algorithms, we thank these authors for their fantastic and efficient codebase:

• UNet: https://github.com/zhixuhao/unet
• FCN & DRN: https://github.com/jhoffman/cycada_release/tree/master/cycada/models