A Two Stage Convolutional Neural Network for Pulmonary Embolism Detection From CTPA Images

By Xin Yang, Yi Lin, Jianchao Su, Xiang Wang, Xiang Li, Jingen Lin, Kwang-Ting Cheng

Introduction

This is the official repo for "A Two Stage Convolutional Neural Network for Pulmonary Embolism Detection From CTPA Images". For more details please refer to our paper. Please cite the paper in your publications if you find the source code useful to your research.

Citing Our Paper

@article{yang2019two,
  title={A Two-Stage Convolutional Neural Network for Pulmonary Embolism Detection From CTPA Images},
  author={Yang, Xin and Lin, Yi and Su, Jianchao and Wang, Xiang and Li, Xiang and Lin, Jingen and Cheng, Kwang-Ting},
  journal={IEEE Access},
  volume={7},
  pages={84849--84857},
  year={2019},
  publisher={IEEE}
}

Requirements

python 2.7
pytorch >= 0.4.0

Usage

Clone the repository

    $ git clone git@github.com:hust-linyi/A-Two-Stage-Convolutional-Neural-Network-for-Pulmonary-Embolism-Detection-From-CTPA-Images.git

Preparation:

Download the FUMPE dataset

And our mantual labels of PE Challenge test set

For stage 1:

    cd stage1

Train

0. Generate *.csv file for groundtruth, refer to ./preprocess/get_3D_label.py. For preprocess, please refer to ./preprocess/preprocess.py

1. In ./stage1/detector/ folder and run:

   train.sh

Test

In ./detector/ folder and run:

    test.sh

For stage 2:

    cd stage2

Train

0. Generate new *.csv file for groundtruth, refer to ./prepare_csv.py

   python classification.py --test 0

Test

    python classification.py --test 1

Notes

For the consideration of patient privacy, the PE129 dataset is not released.

You can download the trained model of stage1 and stage2.