c3d question

[nadluru]

Dear ANTs community members,

I am trying the following c3d commands.

The first one listed below (slightly adapted from c3d documentation) works as expected that it produces PNG with segmentation overlaid on T1w.

c3d T1w.nrrd -stretch 2% 98% 0 255 -clip 0 255 \
tissue_labels.nrrd \
-foreach -slice z 50% -endfor \
-oli tissue_labels.txt 0.25 -type uchar -omc overlaid_approach1.png

I then tried the following adaption which conceptually feels like equivalent but is producing a PNG with only the T1w and not with the segmentation overlaid.

c3d T1w.nrrd -stretch 2% 98% 0 255 -clip 0 255 -slice z 50% \
tissue_labels.nrrd -slice z 50% \
-oli tissue_labels.txt 0.25 -type uchar -omc overlaid_approach2.png

The main difference between the two approaches is in where the -slice z 50% operation is performed. I uploaded the data here ants_forum_c3d.zip. Could you kindly share your thoughts on why there would be a difference between the two approaches?

I realize c3d might not be officially part of this GitHub discussion but I could not find "discussion" feature on the c3d GitHub and I thought the community here would have the most experience with that tool. Thank you so much for your consideration of my request.

Sincerely,
Nagesh

[cookpa]

There is an ITK-SNAP users list

https://groups.google.com/g/itksnap-users

where you can also ask about c3d.

Looking at the command, my guess is that the second one works differently because it doesn't have -foreach -slice z 50% -endfor. This changes the order of images on the stack when -oli is called.

[nadluru]

Thank you for pointing to that users list, Philip!

My understanding is that the order of the images on the stack would match the order in which they appear in the command, and -foreach -slice z 50% -endfor also preserves that order. And even though the order specification is the same in both the commands, as you guessed, somehow the order is being changed for the second approach but would like to understand why? I will post the question to the itksnap-users google group. Thanks again, Philip