open-discussion > Representing closed loops in SWC
Showing 1-1 of 1 posts
Mar 7, 2018 07:03 PM | esteves5 - University of Toronto
Representing closed loops in SWC
Hello!
This topic was brought up a couple years ago but I havn't seen any further discussion about it.
It is possible for SWC files to record closed loop structures given that each branch is supposed to only have one parent? I can see in my automatic reconstructions using Vaa3d many branches that contact each other.
In one instance I saw a closed loop but wasn't able to find this point in the SWC file to see how it was recorded. Often when the branches in the image contact each other the corresponding trace shows the skeleton branches in very close proximity but technically are not connect so the closed loop structure is not documented in the SWC file.
I am interested in how closed loops can be detected in SWC files since these structures often occur during development before excess branches are pruned.
I was thinking that using the x, y, z co-ordinates could capture these features. If you cross-reference points where 2 co-ordinates are the same and the third co-ordinate is constrained to only vary by a small distance then those 2 point would essentially be very close together, perhaps at a point of contact forming a closed loop.
What do you think?
Cheers,
Sam
This topic was brought up a couple years ago but I havn't seen any further discussion about it.
It is possible for SWC files to record closed loop structures given that each branch is supposed to only have one parent? I can see in my automatic reconstructions using Vaa3d many branches that contact each other.
In one instance I saw a closed loop but wasn't able to find this point in the SWC file to see how it was recorded. Often when the branches in the image contact each other the corresponding trace shows the skeleton branches in very close proximity but technically are not connect so the closed loop structure is not documented in the SWC file.
I am interested in how closed loops can be detected in SWC files since these structures often occur during development before excess branches are pruned.
I was thinking that using the x, y, z co-ordinates could capture these features. If you cross-reference points where 2 co-ordinates are the same and the third co-ordinate is constrained to only vary by a small distance then those 2 point would essentially be very close together, perhaps at a point of contact forming a closed loop.
What do you think?
Cheers,
Sam