open-discussion
8 Subscribers 
open-discussion > Closed loops in automatic neuron reconstruct.
Showing 1-2 of 2 posts
Mar 10, 2015 03:03 PM | Miroslav Radojevic
Closed loops in automatic neuron reconstruct.
How would be your experiences regarding closed loop traces in
neuron reconstruction? I am curious about an expert opinion. Swc
standard sort of forbids closing the loops (the definition of the
format that you always refer to the parent node). I've also found
statements in the literature saying that it is illegal to have
those.
What happens on the engineering side in this case? Perhaps there is a standard method on embedding closed loop traces in swc structure? In reality the closed loops exist (happens with neuron structure especially in 2d, and it is not fair to discard those traces, on the other hand they are treated as an error in metrics that take into account the structure) - is there any standard approach? I tried finding a way to still incorporate those into the reconstruction however - it is not easy, and I might be doing something that is not even feasible (it looks so).
What happens on the engineering side in this case? Perhaps there is a standard method on embedding closed loop traces in swc structure? In reality the closed loops exist (happens with neuron structure especially in 2d, and it is not fair to discard those traces, on the other hand they are treated as an error in metrics that take into account the structure) - is there any standard approach? I tried finding a way to still incorporate those into the reconstruction however - it is not easy, and I might be doing something that is not even feasible (it looks so).
Mar 10, 2015 03:03 PM | Hanchuan Peng
RE: Closed loops in automatic neuron reconstruct.