# Learning from ambiguous labels for lung nodule malignancy prediction

This repo contains the official implementation of our paper: Learning from ambiguous labels for lung nodule malignancy prediction, which proposes a multi-view 'divide-and-rule' (MV-DAR) model to learn from both reliable and ambiguous annotations for lung nodule malignancy prediction on chest CT scans. The implementation of DAR model is released.

<p align="center"><img src="https://raw.githubusercontent.com/Merrical/DAR/master/MVDAR_overview.png" width="90%"></p>

#### [Paper on IEEE](https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9705525)

#### [Paper on arXiv](https://arxiv.org/pdf/2104.11436.pdf)

### Requirements

This repo was tested with Ubuntu 20.04.4 LTS, Python 3.8, PyTorch 1.9.0, and CUDA 10.1.

We suggest using virtual env to configure the experimental environment.

1. Clone this repo:

```bash

git clone https://github.com/Merrical/DAR.git

```

2. Create experimental environment using virtual env:

```bash

virtualenv .env --python=3.8 # create

source .env/bin/activate # activate

pip install -r requirements.txt

```

### Bibtex

```

@article{liao2022learning,

  title={Learning from ambiguous labels for lung nodule malignancy prediction},

  author={Liao, Zehui and Xie, Yutong and Hu, Shishuai and Xia, Yong},

  journal={IEEE Transactions on Medical Imaging},

  year={2022},

  publisher={IEEE}

}

```

### Contact Us

If you have any questions, please contact us ( merrical@mail.nwpu.edu.cn ).