This is an example of the CT imaging is used to segment Lung Lobe.
The following dependencies are needed:
- numpy >= 1.11.1
- SimpleITK >=1.0.1
- tensorflow-gpu ==1.14.0
- pandas >=0.20.1
- scikit-learn >= 0.17.1
1、Preprocess
- download CT Lung Lobe data:https://github.com/deep-voxel/automatic_pulmonary_lobe_segmentation_using_deep_learning
- generate patch(144,112,112) image and mask image:run the data3dprepare.py
2、Lung Lobe Segmentation
- the Vnet3d model can find here:https://github.com/junqiangchen/VNetFamily
- train and predict in the script of train.py and inference.py
-
test dataset segmentation result:left is groundtruth,right is predict
-
more detail can follow my WeChat Public article.
- https://github.com/junqiangchen
- email: [email protected]
- Contact: junqiangChen
- WeChat Number: 1207173174
- WeChat Public number: 最新医学影像技术