# ISBI2018_PETCT_Segmentation

**Repository Path**: peakb_admin/ISBI2018_PETCT_Segmentation

## Basic Information

- **Project Name**: ISBI2018_PETCT_Segmentation
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: MIT
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2025-07-29
- **Last Updated**: 2025-07-29

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# ISBI2018_PETCT_Segmentation

This repository contains the code (in TensorFlow) for "[3D fully convolutional networks for co-segmentation of tumors on PET-CT images](https://ieeexplore.ieee.org/abstract/document/8363561/)" paper (ISBI 2018). Compared to the previous semi-automated methods, this method is highly automated without manually user-defined seeds. 

**UPDATED**

1. Uploaded the DFCN-CoSeg training and testing code for our extended work published in
https://aapm.onlinelibrary.wiley.com/doi/abs/10.1002/mp.13331 (MP2018), which provided much details
compared to the ISBI2018 paper. 

2. Uploaded our previous trained models for `CT-Only`, `PET-Only` and `DFCN-CoSeg` networks studied in 
MP2018. The models can be downloaded in 
(1) BaiduYun (https://pan.baidu.com/s/1tCsjfuckkU9IH8O4xewsRQ Password: tfkt), 
or (2) https://app.box.com/s/9r7zxfcs5y9kr5woa1bze8v2lgz48ryv.

3. As for now, I cannot install the outdated `tensorflow_gpu==1.4` in my working `Ubuntu 20.04`, 
so I uploaded two cases of PET-CT images and the testing code using `tensorflow_gpu==2.3`, 
interested readers can check the `test.sh` script.
**Please note that we just use the `tensorflow_gpu==2.3` in the testing code, not for training.**

4. **With regarding to the PET SUV computation, please refer to the NCI-QIICR project (http://qiicr.org/tool/PETDICOM/), they have introduced an implementation as an extension for the open source 3D Slicer software (https://www.slicer.org/).**

## CT/PET Segmentation Results on One Patient

### 1. CT image

<img align="center" src="https://github.com/zhongzisha/ISBI2018_PETCT_Segmentation/raw/master/CT.PNG">

### 2. PET_SUV image

<img align="center" src="https://github.com/zhongzisha/ISBI2018_PETCT_Segmentation/raw/master/PET_SUV.PNG">

### 3. Ground Truth Segmentation on CT image

<img align="center" src="https://github.com/zhongzisha/ISBI2018_PETCT_Segmentation/blob/master/CT_Ground%20Truth.PNG">

### 4. Ground Truth Segmentation on PET_SUV image

<img align="center" src="https://github.com/zhongzisha/ISBI2018_PETCT_Segmentation/blob/master/PET_Ground%20Truth.PNG">

### 5. Prediction on CT image

<img align="center" src="https://github.com/zhongzisha/ISBI2018_PETCT_Segmentation/blob/master/Prediction_CT.PNG">

### 6. Prediction on PET_SUV image

<img align="center" src="https://github.com/zhongzisha/ISBI2018_PETCT_Segmentation/blob/master/Prediction_PET.PNG">

### 7. Wrong Predictions on CT image

<img align="center" src="https://github.com/zhongzisha/ISBI2018_PETCT_Segmentation/blob/master/Prediction_CT_Wrong.PNG">

### 8. Wrong Predictions on PET_SUV image

<img align="center" src="https://github.com/zhongzisha/ISBI2018_PETCT_Segmentation/blob/master/Prediction_PET_Wrong.PNG">

## Dependencies

- [Python2.7](https://www.python.org/downloads/)
- [TensorFlow(1.4.0+)](http://www.tensorflow.org)
- [DLTK](https://dltk.github.io/)
- other libraries

## Citation

If you find this useful, please cite our work as follows:

```
@INPROCEEDINGS{zszhong2018isbi_petct,
  author={Z. Zhong and Y. Kim and L. Zhou and K. Plichta and B. Allen and J. Buatti and X. Wu},
  booktitle={2018 IEEE 15th International Symposium on Biomedical Imaging (ISBI 2018)},
  title={3D fully convolutional networks for co-segmentation of tumors on PET-CT images},
  year={2018},
  volume={},
  number={},
  pages={228-231},
  keywords={Biomedical imaging;Computed tomography;Image segmentation;Lung;Three-dimensional displays;Tumors;co-segmentation;deep learning;fully convolutional networks;image segmentation;lung tumor segmentation},
  doi={10.1109/ISBI.2018.8363561},
  ISSN={},
  month={April},
}

@article{zszhong2018mp_petct,
  author = {Zhong, Zisha and Kim, Yusung and Plichta, Kristin and Allen, Bryan G. and Zhou, Leixin and Buatti, John and Wu, Xiaodong},
  title = {Simultaneous cosegmentation of tumors in PET-CT images using deep fully convolutional networks},
  journal = {Medical Physics},
  volume = {46},
  number = {2},
  pages = {619-633},
  keywords = {cosegmentation, deep learning, nonsmall cell lung cancer (NSCLC), tumor contouring},
  doi = {10.1002/mp.13331},
  url = {https://aapm.onlinelibrary.wiley.com/doi/abs/10.1002/mp.13331},
  eprint = {https://aapm.onlinelibrary.wiley.com/doi/pdf/10.1002/mp.13331},
  year = {2019}
}

```

## Contacts
zhongzisha@outlook.com

Any discussions or concerns are welcomed!