This repository contains our GUI application for automatic spine segmentation and refinement.
We provide an executable ready to run in Windows 10. To have short segmentation times, we recommend running our application on a system with an NVIDIA GPU with at least 8 GB of VRAM. Otherwise, the segmentation process will be done on CPU and will be slower. To use this executable:
DeepSpineNetTool.zip
from https://bit.ly/3usWxyvDeepSpineNetTool.bat
If you have a compatible GPU, please ensure that your GPU drivers are up-to-date.
Our application can also be installed in any operating system that supports its dependencies. It requires Python 3.6.8 or later and CUDA 10.1 (to enable the use of GPUs to greatly reduce the automatic segmentation time. It can be used without CUDA on CPU, with a longer execution time.).
git clone https://gitfront.io/r/user-4306573/be116855b22f779ae17fb981f89fbd138ac27133/DeepSpineNet-GUI.git
pip install -r requirements.txt.
models.zip
) from https://bit.ly/3usWxyvAfter the previous steps, folder structure should be:
DeepSpineNet-GUI
app
models
M1
M2
M3
To present our application functionality, we provide a sample project. It can be downloaded from: https://bit.ly/3usWxyv (sample.scn
)
To start the tool, use the following command (if using the executable: Open DeepSpineNetTool.bat
):
python main.py
Once the main window appears, in the upper menu, open the sample project with Scene
> Load
and accept the prompt message.
Locate the sample project and open it.
Scene Manager
pane, for example, you can choose ROI_RAW_test6.tif
Image
> Viewers
and click on Basic Image 3D Viewer
ROI_RAW_test6.tif
in the Scene Manager
paneSegmentation
> Deep Learning
and click one of M1
M2
or M3
(If no option shows when hovering over Deep Learning
, the model folder has not been placed correctly, please check the installation instruction)Close when finished
checkbox from the progress prompt and wait for the process to finishScene Manager
pane (ROI_RAW_test6.tif (Seg: u_net3d_deep)
) and can be viewed following the steps from the previous sectionROI_RAW_test6.tif
and ROI_RAW_test6.tif (Seg: u_net3d_deep)
(or ROI_LABEL_test6.tif
if you didn't perform the automatic segmentation) in the Scene Manager
pane.Segmentation
and click on Segmentation Editor
.ROI_RAW_test6.tif (Seg: u_net3d_deep)
(or ROI_LABEL_test6.tif
if you didn't perform the automatic segmentation), from the list. The segmentation editor will open.edition_ROI_RAW_test6.tif (Seg: u_net3d_deep)
), if you want to save the current work and continue it later, next time you will have
to open the editor with the 3 images selected: ROI_RAW_test6.tif
, ROI_RAW_test6.tif (Seg: u_net3d_deep)
and edition_ROI_RAW_test6.tif (Seg: u_net3d_deep)
.The authors gratefully acknowledges the computer resources at Artemisa, funded by the European Union ERDF and Comunitat Valenciana as well as the technical support provided by the Instituto de Física Corpuscular, IFIC (CSIC-UV).
DeepSpineNet GUI is distributed under a Dual License model, depending on its usage. For its non-commercial use, it is released under an open-source license (GPLv3). Please contact us if you are interested in commercial license.