FastMarchingSegmentation Graphical Interface?

Hi Guys,

I want to apply the FastMarchingSegmentation Filter. In the documentation
https://simpleitk.readthedocs.io/en/master/Examples/FastMarchingSegmentation/Documentation.html is mentioned that it’s possible to view the output over time

A graphical interface can be constructed to show the contour propagation over time, enabling a user to select a the desired segmentation.

What kind of graphical interface? It’s there an algorithm/library already implemented for the GUI? I don’t even know where to begin.

Many thanks.

Thank

Hello,

If you have seen them before the SimpleITK Jupyter Notebooks contain some basic GUI elements with slicers and images, which is all that is needed here. The output of the Fast Marching represents the monotonic propagation of the front. You can just a threshold filter to see the segmentation a specific time. Then connect the threshold value or fast marching time to a slider and you have a basic GUI to select the desired fast marching time for a segmentation.

@zivy May be able to point you to a specific notebook to get started from, but I’d start with the thresholding and levelset notebooks.

1 Like

I think I have found the SimpleITK Jupyter Notebook.

http://insightsoftwareconsortium.github.io/SimpleITK-Notebooks/Python_html/300_Segmentation_Overview.html

However I cannot find any module called myshow. I get an unresolved reference.

Found it : https://simpleitk-prototype.readthedocs.io/en/latest/user_guide/myshow.html

Also possibly look at gui.py which includes several UI’s for various display purposes.

3D Slicer has a GUI for fast marching segmentation with live preview in 2D/3D views:

It uses vtkPichonFastMarching class internally and not sitk.FastMarchingImageFilter, but it should be easy to modify the Python script of the segment editor effect to use ITK’s filter if you need that. You can use source code of Watershed effect as an example of using SimpleITK filter for segmentation.

2 Likes